Import a share into Unity Catalog and query on-prem data in place—no copying, no downsampling, results back in seconds. Watch the 7-minute walkthrough of MinIO AIStor Table Sharing connected live to Databricks.

Watch the Demo

👋 Hi there, welcome to DataPro #184 – The AI Breakthrough Is Easy. Proving It Works Is Harder.

AI teams are getting remarkably good at building things that look successful.

The harder question is whether those systems are actually improving the user experience.

We lead this issue with Lorenzo Toni, Lead Data Scientist, AI Product Analytics at Amazon, in conversation with Nilesh Kowadkar, Relationship Lead at Packt, on one of the most overlooked problems in AI product development: how do you measure whether an AI feature is genuinely working?

Using a generalized e-commerce chatbot experiment, Lorenzo shows why the obvious signals can be misleading. Engagement increased and conversations became longer, yet conversion eventually declined. The lesson is bigger than one chatbot: AI experimentation cannot stop at surface-level metrics. Teams need to understand downstream behaviour, account for the inherent randomness of generative AI, and use causal inference alongside experimentation to separate correlation from genuine product impact.

That same shift from impressive AI capabilities to measurable, production-ready systems runs through the rest of the edition.

What Else We’re Tracking

• Google Cloud moves enterprise AI closer to production, with Gemini Enterprise for Financial Services and Legal, tighter agent cost controls, expanded Antigravity access, and continued work on hallucinations and token economics.

• AWS rethinks support operations with GenAI, combining automated video-to-SOP creation, RAG, agentic workflows and ML-based SLA risk prediction to turn fragmented operational knowledge into actionable intelligence.

• Google’s TabFM brings zero-shot predictive ML to BigQuery, using in-context learning to perform classification and regression through SQL without the traditional train, tune and deploy cycle.

• AWS tackles the monitoring problem infrastructure checks cannot see: dashboards that load successfully but contain blank visuals or incorrect numbers. Its Bedrock-powered validation system reduced detection time from as much as 72 hours to under one hour.

• BigQuery Graph is now generally available, bringing graph analytics, GQL, SQL, ML and AI together to uncover connected relationships and provide richer grounding for AI agents and GraphRAG.

• Perplexity open-sources Lily, a purpose-built Rust + Metal inference engine that pushes Qwen3.6-35B-A3B performance on Apple Silicon beyond MLX-LM in its reported benchmarks.

• Code-as-World asks whether AI can model physics, not just pixels, turning real videos into executable MuJoCo worlds through an agentic propose-and-verify loop.

From measuring AI impact and controlling agent costs to predictive ML, graph-grounded agents, local inference, and executable world models, this issue looks beyond what AI can generate to ask the harder engineering question:

Can we make AI measurable, reliable, efficient, and genuinely useful in the real world?

Let’s get into it!

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Live LLM Engineering Masterclass

📅 Saturday, September 12 | 09:30 AM–1:30 PM EDT

Build reliable LLM systems with production evals, RAG, resilient agents, regression testing, observability and cost control.

Save 45% with PRODLLM45

Your AI Experiment Says the Feature Is Working. Your Users May Be Telling You Otherwise

An AI feature can pass an experiment and still make the product worse.

Engagement can rise. Session duration can increase. Users can interact with the feature more frequently. An A/B test can even produce statistically convincing results.

And yet, a few weeks later, conversion starts falling.

Users stop returning.

Trust begins to erode.

That is the measurement problem AI product teams increasingly need to confront.

Traditional digital products generally give users a relatively deterministic experience. Change a button, ranking algorithm, checkout flow, or interface and teams can establish a reasonably clean treatment and control.

Generative AI changes that assumption.

The same feature can produce different answers for different users, and even different answers for the same user. A seemingly insignificant prompt adjustment can alter the length, tone, relevance, accuracy, and usefulness of an interaction.

So what exactly should teams measure?

And more importantly, how do you know whether an AI feature actually caused the outcome you are seeing?

That question was at the centre of a conversation between Nilesh Kowadkar, Relationship Lead at Packt, and Lorenzo Toni, Lead Data Scientist — AI Product Analytics at Amazon, focused on experimentation and causal inference for AI products. Drawing on a generalized e-commerce case study, Lorenzo unpacked why apparently positive metrics can be misleading and why AI teams need to look beyond conventional A/B testing.

Partner Spotlight | Building a startup? Pitch live to 20 investors in one day.

Fundraising shouldn’t mean spending weeks cold-emailing VCs and waiting for replies.

We’re partnering with VC Pitch Conference to give founders a more direct route to investors. On September 17, selected startups can join 20 guaranteed 1:1 investor calls in a single day, with meetings matched by sector, stage, and geography.

With 290+ VCs and angel investors looking across AI, fintech, B2B, and other sectors, it’s an opportunity to put your startup directly in front of investors actively looking for their next deal.

Packt readers get an additional 10% off the early-bird ticket with code PACKT10.

Pitch Your Startup to Investors →

When More Engagement Is Actually a Warning Sign

Consider an e-commerce platform introducing an AI chatbot designed to help customers discover products.

Instead of navigating through search results, users could ask questions, compare prices, describe what they wanted, and receive recommendations conversationally.

A change was made to the prompt powering the chatbot.

Initially, the numbers looked promising.

Engagement increased. Users performed more actions. They also spent longer interacting with the chatbot.

Under a conventional product analytics framework, those signals could easily be interpreted as evidence that the new experience was working.

Then, roughly two weeks later, another pattern emerged.

Chatbot interactions began declining.

More importantly, overall website conversion started falling.

The metric that initially appeared particularly encouraging, time spent with the chatbot, turned out to be negatively correlated with conversion.

Users were not necessarily spending longer because the experience was valuable.

Some were spending longer because the chatbot was taking too long to help them reach a decision.

A ten-minute conversation might look like excellent engagement on a dashboard. To the customer, it might mean:

I could have found this myself in thirty seconds.

That distinction matters.

Because when AI introduces friction instead of removing it, the damage may extend beyond a single failed interaction. Users learn from that experience. The next time they need something, they may bypass the chatbot entirely.

The apparent engagement win can therefore become a trust problem.

How Modern AI Systems Really Find Answers: Build GraphRAG Applications

📅 Saturday, September 19 | 09:30 AM–1:30 PM EDT

Move beyond vector search and build explainable AI with Neo4j, knowledge graphs, Cypher and LLM agents.

Save 35% with MODERNAI35

The Real Problem Wasn’t the Chatbot. It Was What the Chatbot Was Optimizing For.

Digging deeper revealed another issue.

The chatbot was largely recommending top-ranked, popular products.

At first glance, that seems perfectly rational.

Popular products already convert well. Recommend them more frequently and customers should be more likely to buy them.

Except that assumption ignores why someone chooses an AI assistant instead of ordinary search.

A shopper typing pen into a search box probably expects popular pens.

Someone opening an AI chatbot may instead ask:

What is the best gel pen for writing on rough paper?

That is a fundamentally different intent.

The value of an LLM-powered shopping experience comes from its ability to understand those more specific needs and personalize the discovery process, rather than simply reproduce the first page of search results.

This changed the experimentation question.

Instead of asking:

Does the new prompt increase chatbot engagement?

The team needed to ask:

What kind of conversational behaviour actually helps users make better decisions?

Experimenting With Different AI Behaviours

Lorenzo worked with engineering and data teams to test alternative experiences.

The existing prompt became the baseline.

Another version made the chatbot more inquisitive, encouraging it to gather more information about the customer’s needs before recommending something.

A further approach explored what Lorenzo described as alternative “gems”: products that might not dominate the rankings but had strong reviews, good conversion performance, and a closer fit with a customer’s specific requirements.

The experiments also extended across regions using a switchback approach.

Rather than permanently assigning one experience to a group, different versions could be switched on and off across time periods. A user interacting with the system at one point might encounter one LLM behaviour and later encounter another.

This provided another way of understanding how changes to the AI experience affected behaviour while accounting for the dynamic nature of the system.

But product quality was only one side of the equation.

A more sophisticated recommendation experience could require access to a much larger product catalogue and richer context. That means additional memory, infrastructure, retrieval complexity, and ultimately cost.

The winning experiment therefore could not simply be the one with the highest conversion rate.

It had to be statistically meaningful, commercially valuable, and operationally practical.

Build Production-Ready AI Agents : From Workflow to Deployment

📅 Saturday, September 26–Sunday, September 27 | 10:00 AM–2:00 PM EDT
Build production-ready AI agents with n8n, from real-world workflows and tool integrations to testing, security, governance, and deployment.

Save 35% with PACKTAI35

Data Science & ML Research Roundup

⭕ What Google Cloud announced in AI this month – and how it helps you. Google Cloud is pushing AI from experimentation into practical enterprise use. August brought Gemini Enterprise for Financial Services and Legal, stronger agent billing and cost controls, and expanded Antigravity access. Alongside this, Google is tackling two stubborn challenges: AI hallucinations and runaway token costs, while helping teams move prototypes into production and start building agents from the ground up.

⭕ Modernizing and scaling support operations with generative AI on AWS: Support teams are drowning in rising ticket volumes, fragmented knowledge, outdated SOPs, and looming SLA breaches. AWS proposes a GenAI-powered operating model that turns training videos into SOPs, uses RAG and agentic workflows to guide ticket resolution, maps process bottlenecks, and applies ML to predict SLA risk, creating a continuous loop where every resolution makes the next one smarter.

⭕ TabFM adds predictive ML to BigQuery: Predictive ML in BigQuery is getting a major shortcut. TabFM, Google’s new tabular foundation model, uses in-context learning to deliver zero-shot classification and regression with a single SQL query. It removes model training, tuning, deployment, and manual feature engineering while scaling predictions across millions of rows, bringing fast, high-quality predictive analytics directly into BigQuery and agentic workflows.

⭕ How an AWS team detects dashboard content failures at scale using Amazon Bedrock: A dashboard can be technically healthy and still show blank charts or wrong numbers. AWS closed this last-mile monitoring gap with an Amazon Bedrock-powered validation system that visually scans hundreds of dashboards, checks numeric consistency, and alerts owners automatically. By combining LLMs for semantic detection with deterministic logic for precision, the team cut failure detection from up to 72 hours to under one hour.

⭕ BigQuery Graph: Connecting Data and AI at Scale: Enterprise data becomes more valuable when you can understand how everything connects. Now generally available, BigQuery Graph brings native graph analytics directly into the warehouse, combining GQL, SQL, ML and AI without ETL. Teams can uncover fraud networks, model supply chains, resolve identities, and build knowledge graphs that ground AI agents and GraphRAG, even across multi-cloud data without moving it.

⭕ Perplexity Open Sources Lily: A Rust + Metal Inference Engine for Qwen3.6-35B-A3B on Apple Silicon. Running a 35B-parameter model locally on a Mac usually means compromising on speed or relying on general-purpose frameworks. Perplexity’s newly open-sourced Lily takes another route: a specialized Rust + Metal inference engine for Qwen3.6-35B-A3B on Apple Silicon. By optimizing GPU routing, dequantization and attention, Lily delivers 1.23x faster prefill and 1.35x faster decoding than MLX-LM.

⭕ Meet ‘Code-as-World’: An Agentic Loop That Rewrites Real Videos Into Executable MuJoCo Physics Programs. Video models can recreate what a scene looks like, but do they understand how the physical world actually works? Code-as-World takes a different approach, converting real videos into editable MuJoCo physics programs. An agentic propose-and-verify loop reconstructs mass, motion, collisions and gravity, creating verified worlds for training. Its 9B model even outperformed Gemini 3.1 Flash on QuantiPhy.

See you next time!

Stop forcing your data into rigid tables. MongoDB Atlas is the developer data platform built for modern analytics — flexible documents, multi-cloud on AWS, and Atlas Search out of the box. Free tier available on AWS Marketplace.

Start for Free

👋 Hi there, welcome to DataPro #181 – Building the infrastructure behind production AI

AI is entering a new engineering phase. The focus is no longer just on building smarter models, but on creating the infrastructure, storage, and operational systems that make AI reliable at scale. This edition explores the technologies powering production AI, from next-generation storage architectures and autonomous database operations to coding agents, multi-agent systems, and cloud-native AI platforms.

🌟 Community Spotlight

AI Data Storage: Challenges, Capabilities, and Comparative Analysis

As AI workloads continue to scale, storage is emerging as one of the biggest infrastructure bottlenecks. In this deep dive, JuiceFS explores why traditional storage architectures struggle with GPU-intensive training, billions of small files, and high-throughput data access, before comparing the capabilities modern AI storage platforms need to support today’s data pipelines.

🎙️ Expert Perspective

The Reliability Gap Holding AI Agents Back

What separates an impressive AI demo from a production-ready AI system? AI researcher and Developer Advocate Sonam Gupta shares the engineering practices behind building reliable AI agents, explaining why evaluation, observability, guardrails, and system design matter far more than model benchmarks when deploying AI at scale.

⚡ This Week’s Highlights

• Behind the scenes of Google Agent Skills

• Prime Agent introduces a self-improving coding harness

• How LendingTree built a production multi-agent mortgage assistant

• Google’s latest AI infrastructure updates

• Meta Muse Code enters beta

• Mobileye automates enterprise support with AgentCore

• Google’s Borderless Lakehouse expands multi-cloud AI

• AWS explains its MCP bridge architecture

• NVIDIA launches Alpamayo 2 Super for autonomous driving

• Google introduces Database Operations Agents

Happy reading!

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

AI Data Storage: Challenges, Capabilities, and Comparative Analysis

Over the past five years, AI has advanced rapidly and has found applications in a wide range of industries. As a storage company, we’ve had a front-row seat to this expansion, watching more and more AI startups and established players emerge across fields like autonomous driving, computational biology, and quantitative investment. AI workloads have introduced new challenges to the field of data storage. Existing storage solutions are often inadequate to fully meet these demands.

In this article, we’ll deep dive into the storage challenges posed by AI workloads, the critical capabilities required to address them, and a comparative analysis of the leading storage solutions spanning cloud and on-premises environments.

Storage challenges for AI

AI workloads have brought new data patterns:

• High-throughput data access challenges: In AI training pipelines, the growing use of GPUs by enterprises has outpaced the I/O capabilities of underlying storage systems. Enterprises require storage solutions that can provide high-throughput data access to fully leverage the computing power of GPUs. For instance, in smart manufacturing, where high-precision cameras capture images for defect detection models, the training dataset may consist of only 10,000 to 20,000 high-resolution images. However, each image is several gigabytes in size, resulting in a total dataset size in the scale of tens or hundreds of terabytes. If the storage system lacks the required throughput, it becomes a bottleneck during GPU-intensive AI training.

• Managing storage for billions of files: AI use cases need storage solutions that can handle and provide quick access to datasets containing billions of files. For example, in autonomous driving, the training dataset consists of small images, each about several hundred kilobytes in size. A single training set comprises tens of millions of such images, each sized several hundred kilobytes. Each image is treated as an individual file. The total training data amounts to billions or even 10 billion files. This creates a major challenge in effectively managing large numbers of small files.

• Scalable performance for hot data: In areas like quantitative investing, financial market data is much smaller in size compared to computer vision datasets. However, this data must be shared among many research teams, leading to hotspots where disk throughput is fully saturated but still cannot satisfy the application’s needs. This indicates that we need storage solutions that can steadily serve a lot of hot data with high throughput and low latency.

The infrastructure landscape for AI has also changed dramatically.

These days, with cloud computing and Kubernetes getting so popular, more and more AI companies are setting up their data pipelines on Kubernetes-based platforms. Algorithm engineers request resources on the platform, write code in Jupyter Notebook to debug algorithms, use workflow engines like Argo and Airflow to plan data processing workflows, use Fluid to manage datasets, and use BentoML to deploy models as applications. Cloud-native technologies have become a standard consideration when building storage platforms. As cloud computing matures, AI applications are increasingly relying on large-scale distributed clusters. With a significant increase in the number of nodes in these clusters, storage systems face new challenges related to handling concurrent access from tens of thousands of pods within Kubernetes clusters.

The evolving application workloads and computing environments significantly change the landscape for professionals managing the underlying infrastructure and platforms. Existing hardware-software-coupled storage solutions often suffer from several pain points, such as no elasticity, no distributed high availability, and constraints on cluster scalability. Distributed file systems like GlusterFS, CephFS, and those designed for high-performance computing (HPC) such as Lustre, BeeGFS, and GPFS are typically designed for physical machines and bare-metal disks. While they can be deployed as large capacity clusters, they cannot provide elastic capacity and flexible throughput, especially when dealing with storage demands in the order of tens or hundreds of billions of files.

Key capabilities for AI data storage

Considering these challenges, we’ll outline essential storage capabilities that are critical for AI scenarios, helping enterprises make informed decisions when selecting storage products.

POSIX compatibility and data consistency

In the AI/ML domain, POSIX is the most common API for data access. Previous-generation distributed file systems, except HDFS, are also POSIX-compatible, but products on the cloud have not been consistent in terms of their POSIX support:

• Compatibility: Users should not solely rely on the description “POSIX-compatible product” to assess compatibility. Instead, use pjdfstest and the Linux Test Project (LTP) framework for testing. We’ve conducted a POSIX compatibility test of cloud file systems for your reference.

• Strong data consistency guarantee: This is fundamental to ensuring computational correctness. Storage systems have various consistency implementations, with object storage systems often adopting eventual consistency, while file systems typically adhere to strong consistency. Careful evaluation is needed when selecting a storage system.

• User mode or kernel mode: Early developers favored kernel mode due to its potential for optimized I/O operations. However, in recent years, we’ve witnessed a growing number of developers “escaping” from kernel mode for several reasons

The Reliability Gap Holding AI Agents Back

AI researcher and Developer Advocate Sonam Gupta explores the engineering practices that transform AI agents from impressive demos into dependable production systems.

Every AI demo looks impressive.

Until it meets a real user.

For the last two years, the AI industry has celebrated increasingly capable foundation models, autonomous agents, and sophisticated reasoning systems. Every product launch promises more intelligence. Every benchmark claims higher accuracy. Every new release inches us closer to what once felt impossible.

Yet behind the excitement lies a quieter reality that every engineering team eventually encounters.

The hardest part isn’t getting an AI system to work once.

It’s getting it to work consistently, reliably, and at production scale.

That challenge is quickly becoming one of the defining engineering problems of the Generative AI era. As AI applications evolve from experimental prototypes into products that millions depend on, success is no longer measured by how intelligent a model appears in a carefully orchestrated demo. It’s measured by how predictably an entire system performs when confronted with the messy, unpredictable behaviour of real users.

At Packt, we believe these are the conversations that deserve more attention.

That is why Sanjana Gupta, Relationship Lead at Packt, sat down with Sonam Gupta, a data scientist, Developer Advocate at Telnyx, and host of the AI Chronicles podcast, for an in-depth discussion on what it really takes to build reliable AI systems. Bringing together her experience of working closely with AI authors, practitioners, and technical experts across Packt’s learning ecosystem, Sanjana steered the conversation beyond the usual questions about models and benchmarks, focusing instead on the engineering realities that practitioners face every day.

As I listened to their discussion, one thing became immediately clear.

This wasn’t another conversation about the latest LLM.

It was about something far more important.

Register now. Save 30% with EMTAX30

Why do so many AI applications fail after the demo, and what separates production-ready AI systems from prototypes that never earn users’ trust?

Over the next thirty minutes, Sonam unpacked lessons from building AI agents, evaluating LLM workflows, deploying voice applications, and interviewing leaders across companies such as Google DeepMind, Microsoft, AWS, and emerging AI startups. Rather than offering theoretical advice, she shared the engineering mindset that modern AI teams need if they want their systems to perform reliably in the real world.What follows is not simply an interview.

It’s a masterclass on why the future of AI belongs not only to those who build intelligent systems, but to those who build systems people can trust.

Build with Ben. Save 45% with OMEM45

The industry is asking the wrong question.

If you’ve attended an AI conference recently, you’ve probably noticed a familiar pattern.

Every discussion eventually circles back to the same debate.

Which model is better?

GPT or Gemini?

Open-source or closed-source?

Should developers switch frameworks?

Will AI agents replace traditional software?

These questions dominate keynote stages, LinkedIn discussions, and engineering forums. They’re important, certainly, but they also risk distracting us from a much larger engineering challenge.

As Sanjana opened the conversation, she didn’t ask Sonam which model she preferred.

Instead, she began somewhere much more fundamental.

Before discussing AI agents, evaluation frameworks, or production systems, she wanted readers to understand the perspective from which Sonam approaches the AI ecosystem itself.

Data Science & ML Research Roundup

⭕ Behind the scenes: How we build, test, and scale Google Agent Skills: Google has revealed how it built and governs Google Agent Skills, the open-source framework that gives AI coding agents structured domain knowledge to improve accuracy, reduce hallucinations, and follow best practices. The post explains how the project scaled from a Google Cloud Next 2026 initiative into a 15,000+ star GitHub repository, backed by standardized skill design, automated quality checks, continuous evaluations, clear ownership, and internal DevRel Skills that help teams automate and standardize workflows.

⭕ Prime Intellect Releases Prime Agent: An Open-Source RLM Harness Where Sub-Agents Are Function Calls Inside Persistent IPython Kernel: Prime Intellect has open-sourced Prime Agent, a self-improving AI coding harness that replaces fixed tool schemas with a persistent IPython kernel, allowing sub-agents to operate as function calls and continuously refine their own prompts, skills, and memory. The release details its deployable architecture, broad model compatibility, benchmark results, including 95.5% on ARC-AGI-3, and real-world use cases spanning software engineering, GPU optimization, AI research, and long-running autonomous coding tasks.

⭕ How LendingTree built a multi-agent mortgage assistant on Amazon Bedrock: LendingTree has shared how it built a production-ready multi-agent mortgage assistant on Amazon Bedrock to help borrowers navigate complex home financing decisions with personalized, compliant AI guidance. The post details its three-agent architecture, built with LangGraph, MCP, Amazon Bedrock Guardrails, and Knowledge Bases, along with deployment lessons, production metrics, and best practices for scaling secure, multi-agent systems in regulated industries.

⭕ What’s new in AI infrastructure this month: Google has published its latest AI infrastructure and orchestration roundup, highlighting new products, performance optimizations, and deployment guides for building and scaling AI and agentic workloads on Google Cloud. The update covers infrastructure launches such as Managed Lustre, C4N VMs, GKE enhancements, TPU tooling, AI security projects, customer deployments, and research showing how Google is improving AI performance, efficiency, and production readiness across its cloud platform.

⭕ Meta AI Releases Muse Code (Beta): A Terminal Coding Agent Powered by the New Muse Spark 1.2 Model: Meta AI has released Muse Code (beta), a terminal-based coding agent powered by the new Muse Spark 1.2 model, designed to tackle complex, long-running software engineering tasks across large codebases. The announcement outlines its persistent async agent architecture, replay-safe runtime, bundled planning skills, co-training approach, benchmark methodology, and real-world GPU kernel optimization case study, positioning it as a production-ready coding assistant for enterprise development workflows.

⭕ How Mobileye transformed support operations using Amazon Bedrock AgentCore: Mobileye has shared how it transformed its internal support operations with an AI Support Agent built on Amazon Bedrock AgentCore, reducing ticket response times by 90% while exceeding 95% accuracy. The post explains its hybrid production architecture, MCP-powered real-time data access, enterprise governance, and how the successful deployment evolved into a self-service platform that enables teams across the company to build and deploy production-grade AI agents.

⭕ Introducing the borderless Lakehouse: Google has unveiled major enhancements to its borderless Lakehouse, enabling AI agents to securely query, reason over, and act on data across on-premises systems, multiple clouds, and SaaS platforms without moving it. The announcement introduces Iceberg REST catalog federation, zero-copy cross-cloud analytics, Knowledge Catalog for trusted agent context, Gemini Enterprise integration, and cost optimizations that simplify building governed, multi-cloud AI agents while reducing data transfer and token costs.

⭕ How we built an MCP bridge to give our AgentCore-hosted AI agent access to local MCP tools: AWS has shared how it built an MCP bridge that lets cloud-hosted AI agents securely use tools and files running on a user’s local machine. The post explains the architecture behind the bridge, including AgentCore Runtime, browser extensions, WebSockets, native messaging, FastMCP, and stdio, showing how remote agents can work with local spreadsheets and systems without exposing credentials or requiring complex infrastructure.

⭕ NVIDIA Releases Alpamayo 2 Super: A 34B Open Vision-Language-Action Model for Robotaxis and Autonomous Driving Under OpenMDW-1.1: NVIDIA has released Alpamayo 2 Super, a 34B open vision-language-action model designed to help autonomous vehicles handle rare, complex driving situations. The release combines full-surround video understanding, trajectory planning, causal explanations, and meta-actions in one system, while introducing commercially usable weights, strong benchmark results, safety-focused reasoning traces, and tools for accelerating fleet-data annotation and autonomous-driving development.

⭕ Deep dive on new AI-powered database agents: Google has introduced Database Operations Agents, a pair of AI-powered assistants that automate database onboarding, monitoring, troubleshooting, and optimization across Google Cloud. The announcement details how the Database Onboarding Agent simplifies database selection and deployment, while the Database Observability Agent uses Gemini, telemetry, and MCP tools to diagnose issues, recommend remediations, and streamline database management across services such as Cloud SQL, Spanner, AlloyDB, and Bigtable.

See you next time!

Beyond the AI Hype: Building Systems People Can Actually Trust

Everyone is talking about AI agents. But the real challenge isn’t building them—it’s building systems your business can actually trust.

As organizations race to deploy enterprise AI, many are discovering that the biggest obstacles aren’t model performance or prompt engineering. They’re fragmented data, weak governance, poor infrastructure, and a lack of accountability.

Leading this edition is an insightful conversation between Marijn Markus, AI Leader at Capgemini, and Packt’s Relationship Lead Sanjana Gupta, exploring why AI doesn’t have an ethics problem—companies do. From data quality and enterprise readiness to the limits of LLMs and the future of responsible AI, it’s a timely reality check on what separates successful AI adoption from expensive experiments.

Also in this week’s highlights:

• 🔷 Google unveils 13 hands-on demos for the Gemini Enterprise Agent Platform, covering the complete AI agent lifecycle.

• 🔷 Amazon introduces an autonomous business intelligence architecture powered by Bedrock AgentCore and MCP.

• 🔷 Google releases OKF v0.2, bringing trust, provenance, and lifecycle metadata to agent-generated knowledge.

• 🔷 AWS introduces Task-Aware Knowledge Compression (TAKC), reducing enterprise AI context costs by up to 64×.

• 🔷 Google demonstrates how to take AI agents from prototype to production with Agents CLI.

• 🔷 Andrew Ng launches OpenWorker, a local-first AI coworker that delivers completed work instead of chat.

• 🔷 Google Cloud expands Conversational Analytics across BigQuery, Looker, AlloyDB, Cloud SQL, and Spanner with enterprise-grade governance.

Let’s dive in.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Register now. Save 45% with MLTRADE45

Marijn Markus: AI Doesn’t Have an Ethics Problem. Companies Do.

The Capgemini AI Leader joins Packt’s Sanjana Gupta to explain why accountability, data quality, and human decision-making — not algorithms — will determine the future of enterprise AI.

Everyone wants to talk about Artificial Intelligence.

Very few want to talk about accountability.

Over the last two years, boardrooms have rushed to launch AI initiatives, governments have drafted AI regulations, and organizations have published glossy frameworks promising “Responsible AI.”

But according to Marijn Markus, AI Leader at Capgemini, we’re obsessing over the wrong problem.

In a compelling conversation with Sanjana Gupta, Relationship Lead at Packt, Markus challenges some of the biggest assumptions surrounding AI — from ethics and hallucinations to enterprise transformation and the future of work.

His central message is both uncomfortable and refreshing:

AI isn’t exposing technological problems. It’s exposing human ones.

Rather than another conversation about prompt engineering or the latest foundation models, this discussion dives into something far more important: why successful AI adoption depends on organizational maturity, high-quality data, and a willingness to confront uncomfortable truths.

AI Ethics Has Become a Convenient Distraction

Ask any enterprise about responsible AI and you’ll hear familiar phrases.

Bias.

Fairness.

Transparency.

Explainability.

Governance.

These are important topics.

But Markus believes they’re only part of a much bigger conversation.

Businesses are increasingly comfortable discussing AI ethics, while avoiding discussions about business ethics.

That’s a crucial distinction.

When organizations automate jobs, optimize for profit, or deploy systems that affect people’s lives, those decisions are rarely made by algorithms.

They’re made by people.

Yet when something goes wrong, blame quickly shifts to the technology.

As Markus points out, AI has become a convenient lightning rod — allowing organizations to distance themselves from decisions that were always fundamentally human.

The real ethical question isn’t whether an algorithm made a poor recommendation.

It’s who chose to deploy it, accepted its risks, and benefited from its outcomes.

The Curious Lifecycle of Enterprise AI

One of the interview’s most memorable observations comes from Markus’ experience leading AI projects across industries.

He jokes that every AI project follows the same lifecycle.

It starts as Artificial Intelligence during the sales pitch.

During implementation it becomes Machine Learning.

When lawyers become involved, it turns into Statistics.

By the time auditors inspect the system…

“It’s just an Excel sheet.”

The room laughs.

But the point is serious.

Technology doesn’t change.

The language does.

And with every change in terminology, accountability quietly shifts.

Organizations often use labels strategically, but changing the vocabulary doesn’t change the impact those systems have on people’s lives.

That’s why Markus insists these conversations shouldn’t be framed as AI dilemmas.

They’re simply ethical dilemmas.

Hallucinations Aren’t Bugs. They’re Probability.

Few topics dominate AI conversations more than hallucinations.

Markus strips away the mystery with a surprisingly simple analogy.

Large language models are sophisticated prediction engines.

Every response is, in essence, a probability calculation — a highly informed guess based on patterns learned from enormous amounts of data.

“They’re rolling dice,” he explains.

Not random dice.

Extremely well-trained dice.

That distinction matters.

Because it means hallucinations cannot be completely eliminated.

Only reduced.

Guardrails.

Evaluation pipelines.

Human review.

Retrieval systems.

These all lower the probability of failure, but they cannot remove uncertainty entirely.

Ironically, Markus argues, humans operate in much the same way.

People make mistakes.

Forecasts fail.

Judgment isn’t perfect.

The goal isn’t perfection.

It’s understanding acceptable levels of risk and designing systems accordingly.

Build with Ben. Save 40% with OME40

The Biggest AI Problem Isn’t AI

As the conversation shifts toward enterprise transformation, Markus introduces what may be his most important observation.

Companies don’t have AI problems.

They have infrastructure problems.

Many organizations proudly announce ambitious AI roadmaps while relying on fragmented databases, disconnected systems, poorly maintained SharePoint repositories, and years of inconsistent documentation.

Then they wonder why building an enterprise chatbot proves difficult.

“How can you build a company GPT,” Markus asks, “when your data isn’t organized in the first place?”

Artificial Intelligence doesn’t replace weak digital infrastructure.

It exposes it.

The organizations succeeding with AI today aren’t necessarily using better models.

They’re the ones that invested years earlier in digitization, governance, metadata, and clean data pipelines.

AI isn’t creating new weaknesses.

It’s shining a spotlight on existing ones.

We’re Running Out of Human Data

One of the interview’s most fascinating discussions centers on an emerging challenge that receives surprisingly little attention.

Future AI systems need high-quality training data.

But today’s internet increasingly consists of content generated by AI itself.

Articles.

Comments.

Images.

Videos.

Social media posts.

Entire conversations between bots.

As synthetic content grows faster than human-created content, future models risk learning from previous generations of AI instead of authentic human knowledge.

It’s a feedback loop that could gradually reduce the quality of future AI systems.

For Markus, this isn’t a theoretical concern.

It’s one of the industry’s biggest long-term challenges.

The future of AI may depend less on building larger models and more on preserving genuine human knowledge.

The Dangerous Rise of “Magical Thinking”

Perhaps the strongest message from the interview is Markus’ warning against what philosophers call magical thinking.

Organizations encounter a difficult business problem.

Instead of understanding it, they decide AI will somehow solve it.

No diagnosis.

No root-cause analysis.

No understanding of whether AI is even the appropriate tool.

Just hope.

According to Markus, this mindset has fueled countless disappointing AI projects.

Technology should never come before understanding the problem.

Sometimes a sophisticated LLM is the right solution.

Sometimes a forecasting model works better.

Sometimes a simple linear regression provides greater transparency.

And sometimes…

AI isn’t needed at all.

The best engineers don’t force AI into every problem.

They choose the right solution for the right problem.

Living in the Age of Deepfakes

As generative AI becomes capable of creating convincing images, videos, and voices, Markus believes society faces another challenge.

Trust.

The next generation won’t simply need AI literacy.

They’ll need information literacy.

Instead of trusting a single video or viral post, we’ll need to verify information across multiple independent sources.

Truth, Markus argues, survives verification.

Misinformation rarely does.

In an age where almost anything can be fabricated, critical thinking becomes our most valuable technology.

Beyond the Hype

Despite his criticisms, Markus is far from pessimistic about AI.

In fact, he’s remarkably optimistic.

He compares today’s generative AI revolution to the arrival of Google decades ago.

When search engines first appeared, they were considered cutting-edge AI.

Today, nobody thinks of Google as AI.

It’s simply part of everyday life.

That’s what successful technology does.

It becomes invisible.

Generative AI, he believes, will likely follow the same path.

The companies that succeed won’t be the ones making the loudest announcements or posting the most AI content on LinkedIn.

They’ll be the organizations quietly solving real problems with thoughtful engineering, quality data, and responsible decision-making.

As the conversation draws to a close, Markus leaves viewers with a reminder that perfectly captures the spirit of the discussion:

“Let’s focus on technology changing people’s lives, rather than just changing the content of internet feeds.”

In a world captivated by AI hype, it’s a refreshingly human perspective — and perhaps the most important lesson of all.

Data Science & ML Research Roundup

🔷 13 demos on Gemini Enterprise Agent Platform: Google has unveiled 13 hands-on demos for its Gemini Enterprise Agent Platform, showcasing the complete AI agent lifecycle from development to production. Built on the Agent Development Kit (ADK), the tutorials cover agent creation, MCP integration, human-in-the-loop workflows, scalable deployment, governance with built-in security, cross-framework orchestration, and continuous evaluation. With the new Agents CLI, developers can scaffold, deploy, monitor, and optimize enterprise-ready agents directly from their coding environment.

🔷 Generate Autonomous Business Insights with AI Agent and MCP Servers: Amazon has introduced a reference architecture for autonomous business intelligence powered by Amazon Bedrock AgentCore, enabling organizations to query enterprise data in natural language without custom integration code. Using pre-built MCP connectors, unified governance, persistent memory, and a semantic data layer, the platform orchestrates insights across IoT, ERP, analytics, and operational systems. The configuration-first approach simplifies secure, scalable multi-agent deployments while reducing data silos and accelerating enterprise decision-making.

🔷 OKF v0.2 adds trust signals: Google has released Open Knowledge Format (OKF) v0.2, extending its open standard for AI agent knowledge sharing with new metadata for provenance, trust, freshness, lifecycle, and attestation. Designed for agent-generated knowledge at scale, the update lets AI systems verify who created information, when it was validated, whether it’s current, and how it was produced—without sacrificing OKF’s lightweight, vendor-neutral design. All additions remain optional, preserving backward compatibility while enabling more trustworthy multi-agent workflows.

🔷 Beyond RAG: Task-aware knowledge compression for enterprise AI on AWS: AWS has introduced Task-Aware Knowledge Compression (TAKC), a technique that goes beyond traditional RAG by pre-compressing enterprise knowledge into task-specific representations for faster, cheaper AI inference. Supporting multiple compression tiers, TAKC preserves cross-document relationships while reducing token usage by 8x–64x. Built on Amazon Bedrock and serverless AWS services, the open-source reference architecture enables scalable, cost-efficient analysis for complex enterprise workloads such as financial due diligence and regulatory compliance.

🔷 Automate agent development lifecycles with Gemini Enterprise: Google has published a hands-on guide to building production-ready AI agents with the Gemini Enterprise Agent Platform and Agents CLI. The tutorial walks developers through the full agent lifecycle—from scaffolding and deterministic tool creation to secure deployment, governance, evaluation, and publishing—entirely within their coding environment. Using an Industry Watch agent as an example, it demonstrates how enterprise AI can combine live data, deterministic workflows, and built-in security to move beyond prototypes into production.

🔷 Andrew Ng Just Released OpenWorker: An Open-Source, Local-First Desktop AI Coworker That Returns Finished Deliverables Instead of Chat. Andrew Ng has released OpenWorker, an open-source, local-first desktop AI coworker that completes tasks instead of generating chat responses. Running entirely on the user’s machine, it combines a Python FastAPI agent, Tauri desktop app, and flexible model routing to produce finished deliverables while keeping data private. Built-in permission controls, prompt-injection safeguards, and support for local or cloud LLMs make OpenWorker a secure, enterprise-friendly AI productivity assistant.

🔷 Conversational Analytics in Google Data Cloud in Q326: Google Cloud has expanded Conversational Analytics across its enterprise data ecosystem, bringing natural language querying to BigQuery, Looker, AlloyDB, Cloud SQL, and Spanner. With support for multi-cloud data, MCP tools, and Gemini Enterprise integration, organizations can securely query business data using AI. The platform also introduces enterprise-grade governance, including row-level access controls, CMEK, VPC, data residency, and HIPAA compliance, enabling trusted, large-scale AI-powered analytics.

See you next time!

For the past two years, AI conversations have revolved around models.

Today, the conversation is shifting to something far more important: the engineering foundations that make those models reliable in production.

As organizations move beyond experimentation, success depends less on choosing the latest LLM and more on building the systems around it. Reliable data pipelines, trustworthy retrieval, rigorous evaluation, secure development workflows, efficient AI agents, and optimized inference are becoming the capabilities that separate impressive demos from production-ready AI.

We begin today’s edition with an expert perspective from Nikola Ilic, founder of Data Mozart, Microsoft Data Platform MVP, Microsoft Certified Trainer, and O’Reilly instructor. In his insightful article, Nikola explains why data engineering is the real foundation of useful LLM applications, and why the quality of your AI system is ultimately determined by the quality of the data workflows behind it.

Also in today’s highlights:

• Anthropic brings multi-agent security reviews to Claude Code with its new security plugin.

• Amazon introduces Self-Distilled Reasoning, making it easier to fine-tune reasoning models without sacrificing general capabilities.

• Google rethinks AI evaluation with Discovery Bench, revealing why traditional benchmarks often hide an agent’s true strengths and weaknesses.

• Cisco Foundation AI open-sources Antares, compact security models that rival much larger systems for vulnerability localization.

• monday.com shares how it deploys AI teammates at enterprise scale using Amazon Bedrock.

• Discover 11 practical principles for building faster, cheaper, and more token-efficient AI coding workflows.

• Google’s Gemini 3.6 Flash, Flash-Lite, and Flash Cyber push AI agents further with lower costs, improved efficiency, and production-ready performance.

Whether you’re building enterprise copilots, RAG systems, AI coding assistants, or autonomous agents, today’s edition explores the engineering practices that turn powerful models into AI systems people can depend on.

🎙️ This Weekend with Packt

Register Now

Happy reading, and we’ll see you in the workshop!

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Why Data Engineering Is the Real Foundation of Useful LLM Applications

The engineering decisions behind every trustworthy LLM application.

Everyone talks about choosing the right LLM. Far fewer people talk about the engineering decisions that determine whether the model ever sees the right information in the first place.

Written by Nikola Ilic

Nikola Ilic is the founder of Data Mozart, a Microsoft Data Platform MVP, Microsoft Certified Trainer, Pluralsight author, O’Reilly instructor, blogger, and international speaker. Known for helping practitioners “make music from data,” Nikola brings deep expertise in Microsoft Fabric, Power BI, SQL Server, analytics engineering, and modern data architecture. Through his writing, training, and hands-on workshops, he helps data professionals design production-ready data platforms that power modern analytics and AI applications.

Most conversations about LLM applications begin with the model.

Which one should we use?

How large is it?

How fast is it?

How much does inference cost?

Those are important questions.

But in production systems — especially Retrieval-Augmented Generation (RAG) applications — they’re rarely the questions that determine whether the application succeeds.

The data workflow does.

The quality of an LLM application is ultimately determined by the quality of the data pipeline behind it. Long before a prompt reaches the model, a series of engineering decisions has already shaped what the model can — and cannot — know.

That’s where reliable AI systems are really built.

You’re Not Connecting Documents to an LLM — You’re Engineering Knowledge

When we build a RAG system or an enterprise knowledge assistant, we are not simply connecting documents to an LLM.

We are building a pipeline that decides what the model is allowed to see.

Before a user asks a single question, source documents have already been ingested, parsed, cleaned, transformed, split into chunks, enriched with metadata, embedded, indexed, retrieved, assembled into context, and eventually evaluated.

Every one of those stages influences the final answer.

If any one of them is weak, the application becomes weaker.

That is why data engineering deserves far more attention in conversations about building LLM applications.

👉 Reserve your seat today and use code DATAPRO35 to save an exclusive 35%

Why Good Models Still Produce Bad Answers

Many teams assume that when an LLM gives an incorrect answer, the model must be at fault.

In reality, the problem often starts much earlier.

Perhaps the right document never entered the system.

A PDF parser may have dropped an important table.

Chunks may have split critical context in half.

Metadata may be incomplete, preventing accurate filtering.

The vector database may still contain stale embeddings from an earlier version of the content.

None of these are model problems.

They’re data engineering problems.

The model can only answer from the evidence it receives. If the evidence is incomplete, poorly structured, or outdated, the answer will be too — no matter how advanced the model is.

This is why improving prompts alone rarely fixes production RAG systems.

The biggest improvements almost always happen upstream.

The Engineering Questions That Actually Matter

Building reliable LLM applications means asking practical engineering questions, not just AI questions.

Questions such as:

• What documents should enter the system?

• How do we preserve structure from PDFs, Markdown, HTML, CSV, JSON, and APIs?

• What chunk size and overlap actually improve retrieval quality?

• Which metadata fields support filtering, governance, lineage, and refresh?

• How do we avoid duplicating vectors every time an ingestion pipeline runs?

• How do we measure whether retrieval quality is improving — or quietly getting worse?

These decisions directly affect cost.

Latency.

Reliability.

Compliance.

Maintainability.

And ultimately, user trust.

They are the engineering foundations that separate experimental AI from production AI.

The Best LLM Systems Are Engineered Data Products

The strongest LLM systems I see are not sophisticated prompt wrappers.

They are engineered data products.

They have repeatable ingestion pipelines.

Inspectable chunks.

Meaningful metadata.

Reliable embedding workflows.

Idempotent vector updates.

Clear source attribution.

Evaluation datasets that expose retrieval failures before users ever encounter them.

They are designed to evolve as new data arrives while remaining reliable, explainable, and observable.

Those capabilities don’t come from choosing a different model.

They come from building better data workflows.

That’s the difference between an impressive demo and a production-ready AI application.

The Skill That’s Becoming Essential for Every AI Builder

As organizations move beyond AI experimentation and begin deploying LLM applications across the enterprise, expectations are changing.

It’s no longer enough to know how to call an API or write effective prompts.

Teams increasingly need engineers who understand the complete lifecycle of knowledge pipelines — from raw documents to trustworthy retrieval.

Whether you’re building enterprise search, AI assistants, internal copilots, or knowledge-driven applications, designing reliable data workflows is rapidly becoming one of the highest-value skills in modern AI engineering.

The engineers who understand these workflows will be the ones building the next generation of production-ready AI systems.

👉 Reserve your seat today and use code DATAPRO35 to save an exclusive 35%.

Build the Complete Workflow — End to End

If these challenges sound familiar, that’s exactly what we’ll be tackling in my upcoming live Packt workshop, Designing Data Engineering Workflows for LLM Applications.

This isn’t a slide-heavy overview or another discussion about prompt engineering.

It’s a hands-on session where we’ll build a complete, production-oriented workflow — from raw documents to a working Retrieval-Augmented Generation pipeline.

Together we’ll cover:

• Document ingestion from multiple sources

• Parsing and preserving document structure

• Chunking strategies that improve retrieval quality

• Embeddings and vector database design

• Metadata for governance, filtering, and lineage

• Building Retrieval-Augmented Generation (RAG) workflows

• Evaluating and improving retrieval performance using practical techniques

By the end of the workshop, you’ll have a much clearer understanding of how reliable LLM applications are engineered — and more importantly, how to build those foundations yourself.

Don’t Wait Until Your RAG System Starts Hallucinating

One of the most common patterns I see is that teams only begin thinking seriously about data engineering after users report incorrect answers.

By then, the damage has already been done.

The model wasn’t the real problem.

The pipeline was.

Building trustworthy AI starts long before the prompt reaches an LLM.

It starts with designing the data workflows that determine what the model knows, how it retrieves information, and whether users can trust the answers it generates.

If production-ready LLM applications are part of your roadmap this year, this workshop is designed to give you the practical engineering mindset and hands-on techniques needed to build them with confidence.

Seats are limited to ensure an interactive, hands-on learning experience, and registrations are filling quickly.

If you’re ready to move beyond AI demos and start building reliable, production-grade LLM applications, I’d love to have you join us.

👉 Reserve your seat today and use code DATAPRO35 to save an exclusive 35%.

📚 Weekend Read: Hot Off the Press

Data Science & ML Research Roundup

✅ Anthropic Releases Claude Security Plugin for Claude Code in Beta: A Multi-Agent Vulnerability Scanner That Runs in Your Terminal. Anthropic has introduced the Claude Security plugin for Claude Code in beta, bringing multi-agent security reviews directly into the developer workflow. Instead of simply flagging vulnerabilities, it can scan an entire repository or recent code changes, verify findings through an independent three-agent review process, and generate reviewable patch files without modifying your code automatically. Running locally within your Claude Code session, the plugin complements existing SAST tools by combining AI-driven analysis with transparent verification, giving developers a practical way to identify, validate, and remediate vulnerabilities before code reaches production.

✅ Exploring self-distilled reasoning for supervised fine-tuning with Amazon Nova: Amazon researchers have introduced Self-Distilled Reasoning (SDR), a fine-tuning technique that preserves reasoning capabilities in Amazon Nova 2 models without requiring expensive chain-of-thought annotations. By generating reasoning traces from the base model and using them during supervised fine-tuning, SDR improves domain-specific performance while preventing catastrophic forgetting. Compared with model merging, it retains general reasoning, math, and coding abilities more effectively, offering a practical, low-cost path to customizing reasoning models.

✅ Evaluate agent performance: Traditional AI benchmarks often reduce performance to a single pass-or-fail score, but Google researchers argue this masks where agents truly succeed or fail. Introducing Discovery Bench, an information theory-based evaluation framework, they measure how increasing query ambiguity affects AI retrieval performance. By systematically varying query specificity, the approach exposes hidden failure points, uncovers flaws in benchmark quality, and provides a richer map of agent capabilities—helping developers improve retrieval systems beyond static benchmark scores.

✅ Cisco Foundation AI Releases Antares: 350M and 1B Open-Weight Models That Localize Known Vulnerabilities Inside Real Codebases. Cisco Foundation AI has open-sourced Antares, a family of compact security language models purpose-built for vulnerability localization—identifying vulnerable files in large code repositories from a vulnerability description. Despite its small size, the 1B-parameter model rivals frontier AI systems, outperforming much larger open models while approaching GPT-5.5 on Cisco’s new VLoc Bench. By combining supervised fine-tuning with reinforcement learning, Antares delivers accurate, low-cost vulnerability triage, highlighting how task-specific training can outperform brute-force model scale for practical software security.

✅ AI Teammates: how monday.com runs production AI agents on Amazon Bedrock. monday.com has revealed how it runs AI Teammates in production at enterprise scale using Amazon Bedrock, with AI agents now integrated into engineering workflows alongside human developers. Its multi-agent platform, Sphera, enables agents to pick up tasks, write code, open pull requests, and collaborate through Slack, GitHub, and monday.com, contributing to a 50%+ increase in per-engineer PR throughput. Backed by AWS infrastructure, persistent memory, automated guardrails, and continuous evaluation, the architecture demonstrates how production-grade AI agents can reliably augment software engineering teams at scale.

✅ Guide to AI Tokenomics: Eleven Principles for Token Efficient Software Engineering. Optimizing token usage is becoming an essential skill for developers working with AI coding assistants. Rather than relying on ever-larger context windows, this guide outlines 11 practical strategies—from choosing the right model and using reusable skills to delegating tasks, automating verification, and starting fresh sessions for new topics—to keep AI workflows faster, cheaper, and more accurate. By reducing context bloat and focusing the model’s attention, developers can improve coding quality, lower costs, and build more efficient human-AI collaboration.

✅ Google Releases Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber: A Cheaper, More Token-Efficient Flash Tier Built for Agentic Workloads. Google has expanded its Flash lineup with Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber, targeting developers building high-volume AI agents. Gemini 3.6 Flash delivers stronger coding and reasoning performance while reducing token usage and inference costs, Flash-Lite prioritizes low-latency, cost-efficient workloads, and Flash Cyber is purpose-built for automated vulnerability detection and patching. Together, the releases underscore Google’s focus on making AI agents faster, more affordable, and better suited for production-scale software engineering.

See you next time!

👋 Hi there, welcome to DataPro #178

That shift is becoming increasingly clear. Across the industry, the conversation is moving beyond benchmark scores and bigger models toward production-ready AI, efficient infrastructure, multimodal systems, browser-native inference, and intelligent automation that delivers measurable business value.

We’re leading this edition with an expert perspective from Diogo Alves de Resende, Business Analytics and Data Science Instructor and former Data Scientist at Zalando SE, who explores how modern AI and analytics teams can move beyond experimentation to build practical, scalable solutions that create lasting impact. If you’re building AI products or data-driven applications, this is a must-read.

This week’s highlights

• ⭐ Expert Voice: Diogo Alves de Resende on building practical, production-ready AI systems that deliver real business value.

• Claude on Google Cloud: Deploy frontier AI with managed infrastructure, global endpoints, enterprise security, and agentic workflows.

• Google AlphaEvolve: Google’s AI optimization agent is now generally available, helping engineers discover better algorithms across ML, logistics, finance, and scientific computing.

• Thinking Machines’ Inkling: A 975B open-weight multimodal MoE with controllable reasoning, 1M-token context, and enterprise-ready fine-tuning.

• Mistral Robostral Navigate: An 8B embodied AI model enabling robots to navigate complex environments using just a single RGB camera.

• AWS Computer Vision + MCP: Bringing vision, agents, and the Model Context Protocol together to build intelligent multimodal AI applications.

• Built Technologies + Amazon Bedrock: How AI-powered document intelligence is transforming real estate finance and reducing document review from days to minutes.

• Gemini Omni Flash & Nano Banana 2 Lite: Google’s newest multimodal creative models for high-speed image generation, video editing, and agentic media workflows.

• LiteRT.js: Run .tflite models directly in the browser with WebGPU and WebAssembly for fast, private, on-device AI.

Whether you’re building production-ready GenAI applications, optimizing large-scale AI systems, deploying multimodal models, or exploring the next wave of embodied and browser-native AI, there’s plenty to unpack in this week’s edition.

Grab your favorite beverage, settle in, and dive into the stories shaping the future of data science, machine learning, and AI engineering.

🎙️ This Weekend with Packt

If you’re building LLM applications that need to work beyond the demo, don’t miss our live workshop: Build Reliable GenAI Applications with AI Evals, Observability & Testing.

📅 Saturday, July 18 | 9:30 AM – 1:30 PM EDT

Happy reading, and we’ll see you in the workshop!

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Grab the bundle before the offer ends.

The Next Generation of GenAI Applications Will Be Built Around Workflows

Why production AI is shifting beyond prompts to orchestrated workflows powered by RAG, Python, APIs, evaluations, and guardrails.

Written by Diogo Alves de Resende

For the past two years, much of the conversation around GenAI has focused on prompting.

How do you write better prompts? Which model performs best? How can you make responses sound more intelligent?

Those questions matter, but they miss a larger shift that’s already happening.

The next generation of GenAI applications won’t be evaluated by how well they answer a single question. They’ll be evaluated by whether they can reliably complete an entire workflow from start to finish.

That is a fundamentally different engineering problem.

Register for the Workshop → Build Your Financial AI Analyst

From Answers to Actions

Consider a Financial AI Analyst.

A typical chatbot can answer questions about a company using what it already knows or by retrieving a few relevant document chunks. That might be enough for a demo.

It isn’t enough for a system someone can actually trust.

A production-ready financial assistant needs to do far more than generate text. It needs to gather evidence, perform calculations, validate results, and explain its reasoning before arriving at a conclusion.

A typical workflow might look like this:

• Retrieve the latest annual report.

• Identify the relevant financial statements.

• Extract revenue, cash flow, margins, and debt figures.

• Connect to a live market data API.

• Calculate returns, volatility, valuation ratios, and trends using Python.

• Compare results against previous reporting periods or competitors.

• Generate a clear, source-backed explanation.

Each step depends on a different capability.

RAG retrieves relevant information from enterprise documents. APIs provide live external data. Python performs deterministic calculations that shouldn’t be delegated to an LLM. The language model orchestrates these components, deciding which tools to use, when additional information is required, and how to communicate the final result.

This is what modern AI engineering increasingly looks like.

AI Systems Are Becoming Workflow Engines

Rather than relying on one large prompt and hoping for the best, developers are designing AI applications as structured workflows.

These systems retrieve information, call tools, execute code, validate intermediate results, and adapt their next actions based on what they discover.

The LLM becomes one component within a larger orchestration layer rather than the entire application.

This shift is enabling developers to build AI assistants that are significantly more useful because they can interact with external systems, perform real computations, and produce grounded outputs instead of plausible-sounding guesses.

But it also introduces a new challenge.

Every Additional Step Creates New Failure Points

The more capable an AI workflow becomes, the more opportunities there are for things to go wrong.

The system might retrieve the wrong document.

It could extract an incorrect financial value.

An API might return incomplete data.

A calculation could be performed using outdated inputs.

Or the model might confidently generate a conclusion that isn’t actually supported by the evidence it collected.

These aren’t isolated problems. They’re engineering challenges that emerge whenever multiple tools, data sources, and reasoning steps are combined into a single application.

That’s why modern AI workflows require more than good prompts.

They require evaluations, structured outputs, source verification, guardrails, prompt injection defenses, and mechanisms that make every step observable and testable.

Reliability isn’t something that’s added after deployment. It has to be designed into the workflow from the beginning.

Register for the Workshop → Build Your Financial AI Analyst

Building AI Systems You Can Trust

As enterprises move beyond experimentation, the definition of a successful GenAI application is changing.

It’s no longer enough for a model to produce an impressive answer.

The entire process behind that answer needs to be transparent, repeatable, and reliable enough to support real business decisions.

That’s where workflow-driven AI engineering is headed — and it’s rapidly becoming one of the most valuable skills for data scientists, ML engineers, and AI practitioners building production systems.

Build One Yourself

In my upcoming live workshop, Build Intelligent Assistants with GenAI, Python & AI Tools, we’ll move beyond theory and build a production-ready Financial AI Analyst from scratch.

Together, we’ll build an end-to-end workflow that combines:

• GenAI and modern LLMs

• Python for deterministic financial analysis

• Retrieval-Augmented Generation (RAG)

• Live market data APIs

• Jupyter Notebook

• Cursor

• Lovable

Along the way, we’ll explore how to integrate retrieval, tool use, evaluations, guardrails, and prompt injection defenses into AI workflows that are designed for real-world reliability rather than simple demonstrations.

If you’re looking to move beyond chatbot prototypes and start building production-ready AI assistants, this workshop is designed to give you a practical architecture you can reuse across financial analysis, enterprise copilots, and intelligent business applications.

📅 Live Online Workshop
Build Intelligent Assistants with GenAI, Python & AI Tools
Saturday, July 25 | 7:00 PM–11:00 PM GMT+5

Register for the Workshop → Build Your Financial AI Analyst

Data Science & ML Research Roundup

✅ Claude at scale on Google Cloud: Frontier AI, built for enterprise production: Running frontier AI at enterprise scale is about far more than model quality. This post explores how Claude on Google Cloud combines Anthropic’s advanced reasoning with Google Cloud’s managed infrastructure to help organizations deploy production-ready AI. It covers global low-latency endpoints, built-in security and compliance, data sovereignty, prompt caching, 1M-token context windows, cost optimization, and how the same platform powers the next generation of agentic AI workflows.

✅ Built Technologies builds an AI-powered document intelligence solution on AWS to power agents across real estate finance: Document-heavy industries like real estate are prime candidates for AI transformation. This post explores how Built Technologies uses Amazon Bedrock and the AWS Intelligent Document Processing Accelerator to turn complex real estate documents into production-ready AI intelligence. It shows how scalable document understanding, agentic workflows, and automated extraction are reducing review times from days to minutes while laying the foundation for next-generation AI applications.

✅ Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort. Thinking Machines Lab has unveiled Inkling, its first open-weight foundation model trained from scratch, built for large-scale customization and fine-tuning. This post explores its Mixture-of-Experts architecture, 1M-token context window, multimodal capabilities, controllable reasoning effort, benchmark performance, deployment options, and where it fits in production AI workflows—from voice and vision agents to cost-efficient, domain-specific enterprise applications.

✅ AlphaEvolve is available for everyone: Optimization is becoming one of AI’s most valuable enterprise capabilities. This post explores the general availability of AlphaEvolve on Google Cloud’s Gemini Enterprise Agent Platform, showing how organizations are using AI to optimize algorithms, code, and complex systems across logistics, chip design, genomics, finance, and scientific research. It also explains AlphaEvolve’s four-step optimization workflow and how teams can deploy it to accelerate production-grade AI and engineering innovation.

✅ Agentic vision: Building visual intelligence with Amazon Bedrock and MCP servers. Building AI that can see, reason, and act has traditionally required stitching together multiple tools and APIs. This post explores how AWS combines Computer Vision, Strands Agents, and the Model Context Protocol (MCP) into a unified framework, enabling developers to build intelligent multimodal applications with streamlined integration, secure access, and production-ready image and video understanding powered by Amazon Bedrock, Rekognition, and OpenSearch.

✅ Mistral AI Releases Robostral Navigate: An 8B Model Enabling Robots to Navigate Complex Environments Using a Single RGB Camera. Embodied AI is taking a major step forward with Mistral AI’s Robostral Navigate. This post explores how the 8B vision-language model enables robots to navigate complex real-world environments using only a single RGB camera and natural language instructions. It covers the model’s novel pointing-based navigation, efficient training approach, reinforcement learning, benchmark-leading performance, and practical applications across manufacturing, logistics, hospitality, and autonomous robotics.

✅ Nano Banana 2 Lite and Gemini Omni Flash available: Creative AI workflows are getting faster and more capable with two new additions to Google’s Gemini Enterprise Agent Platform. This post explores Gemini Omni Flash for conversational video generation and editing, alongside Nano Banana 2 Lite, Google’s fastest and most cost-efficient image model. It covers their multimodal capabilities, enterprise governance, price-performance advantages, and how businesses are using them to build next-generation creative and agentic AI applications.

✅ Google Releases LiteRT.js: A JavaScript Binding of LiteRT That Runs .tflite Models in Browsers via WebGPU. Running AI models directly in the browser is becoming faster and more practical. This post explores Google’s new LiteRT.js, a JavaScript binding that brings native LiteRT (.tflite) models to the web using WebAssembly, WebGPU, and experimental WebNN. It covers the architecture, performance gains, model conversion workflow, deployment patterns, and how developers can build private, low-latency, on-device AI applications without relying on cloud inference.

See you next time!

This NINJIO Insights Report dives into the key emotional susceptibilities that make social engineering work and offers concrete steps that your security team can take to equip your workforce to resist cyberattacks.

Download the Guide

👋 Hi there, welcome to DataPro #177 – Special Weekend Edition.

The AI conversation has become increasingly dominated by models, benchmarks, and agents. Every week brings another breakthrough, another framework, or another promise of autonomous intelligence. Yet inside enterprise teams, a far more practical question continues to surface: How do you know whether your AI product is actually delivering value?

Building an AI application is only the beginning. Understanding how users interact with it, where they lose trust, what drives adoption, and whether it improves real business decisions is quickly becoming one of the most valuable skills for data professionals. It’s a discipline where data engineering, analytics, product thinking, SQL, Python, experimentation, and AI all converge.

That’s exactly what we’re exploring in this special weekend edition. In the latest Packt Talks session, hosted by Abhishek Kaushik, Prithvi Shivshankar, Data AI Product Manager at H-E-B, takes us inside the emerging world of AI Product Analytics—showing how modern teams instrument AI products, analyze user journeys, measure product adoption, and use behavioral data to continuously improve AI-powered experiences. Drawing on real enterprise examples, he walks through why successful AI products aren’t built around models alone, but around better decisions.

If you’re building AI applications, working with data products, or simply curious about what happens after an AI product reaches production, this is one read that’s well worth your weekend.

Happy reading!

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Book Your Spot & Save 35%

Why Every AI Product Needs Better Analytics Before Better Models

How SQL, Python, product analytics, and user behavior are becoming the foundation of successful AI products

Every AI product generates outputs.

Far fewer generate outcomes.

As enterprises race to embed generative AI into applications, copilots, and decision-support systems, the conversation has largely centered around models, prompts, reasoning capabilities, and agents. Yet once these products reach production, many organizations encounter a very different challenge.

Are people actually using them?

Do users trust the recommendations?

Where do they abandon the workflow?

More importantly, did the AI improve a business decision—or simply become another dashboard?

That’s exactly what the latest episode of the Packt Talks YouTube series, hosted by Abhishek Kaushik, set out to answer. Prithvi Shivshankar, Data AI Product Manager at H-E-B, walked viewers through the emerging discipline of AI Product Analytics, demonstrating how modern product teams combine SQL, Python, behavioral analytics, experimentation, and business metrics to understand user journeys, evaluate AI performance, and build AI products that people actually trust and use.

Drawing on his experience building enterprise-scale data products at one of the largest independently owned grocery retailers in the United States, Prithvi offered a practical perspective that many AI discussions overlook. The session wasn’t about choosing the best foundation model or writing better prompts. It was about what happens before an AI product is built—and, more importantly, what happens after it ships.

AI Product Analytics Starts with the Business, Not the Model

One of the first ideas Prithvi challenged was the way many AI projects begin.

Too often, teams start with the technology.

A powerful LLM.

An intelligent assistant.

A sophisticated recommendation engine.

Only afterwards do they begin searching for somewhere to use it.

According to Prithvi, successful AI products reverse that process entirely.

Before introducing AI, teams must first understand how decisions are made today. Who makes those decisions? What information do they rely on? Where do they spend most of their time? Which parts of the workflow create friction? Where does decision fatigue begin?

Only after mapping the existing business process should AI enter the conversation. Otherwise, even the most capable model risks becoming another impressive demonstration that never translates into measurable business value.

This simple shift—from starting with technology to starting with workflow—became the foundation for everything that followed during the session.

A Real-World AI Product: Solving Decision Latency in Retail

Rather than relying on theoretical examples, Prithvi grounded the discussion in a production AI product built for supply chain operations.

Managing inventory across a large grocery network is a constant balancing act.

Products move between suppliers, warehouses, transportation fleets, distribution centres, and retail stores before finally reaching customers. Every stage introduces new variables—supplier delays, transportation constraints, seasonal demand, promotions, warehouse capacity, and product shelf life.

Take everyday essentials like milk or bread.

Running out of stock means disappointed customers and lost revenue.

Holding too much inventory means spoilage, shrinkage, and unnecessary operational costs.

The challenge isn’t that organizations lack data.

It’s that analysts often spend hours navigating multiple dashboards, exporting reports, comparing spreadsheets, writing SQL queries, and stitching together information before they can confidently decide whether inventory should be transferred, expedited, reordered, or held.

That’s where the Inventory Decision Assistant comes in.

Its purpose isn’t to replace analysts.

It’s to reduce the time between identifying a problem and taking action.

Throughout the webinar, Prithvi referred to this as reducing decision latency—helping analysts reach the right decision faster by surfacing risks, explaining root causes, and recommending the next best action.

Book Your Seat & Save 35%

AI Doesn’t Create Value by Making Recommendations

Another important theme emerged as the session progressed.

Generating recommendations has become relatively easy.

Generating recommendations that people trust is much harder.

Prithvi explained that an effective AI product needs several layers working together.

It needs to detect meaningful business signals.

Provide enough context to explain why something happened.

Recommend an action.

Allow users to execute that action seamlessly.

Finally, it needs to learn whether that recommendation actually helped.

Without these layers, organizations often mistake model output for business value.

Recommendations alone create very little impact.

Recommendations that users understand, trust, and act upon are what ultimately improve business outcomes.

Mapping the User Journey Before Writing Any AI

One of the most practical parts of the webinar focused on mapping the decision journey.

Using the inventory assistant as an example, Prithvi broke the process into several distinct stages.

Everything begins with a signal—a stock-out risk, declining inventory, or an unexpected demand spike.

Analysts then begin investigating by opening dashboards, querying databases, reviewing historical demand, and comparing operational reports.

Next comes diagnosis.

Is demand increasing unexpectedly?

Has a supplier delayed shipments?

Is inventory sitting too long in a warehouse?

Is there an imbalance between distribution centres?

Only after answering those questions can analysts make a decision.

Transfer inventory.

Expedite shipments.

Adjust purchase orders.

Escalate operational issues.

Finally comes execution, where those decisions translate into actions across the supply chain.

By visualising every step in this journey, teams can identify precisely where AI removes friction—and where simpler automation or better process design may already solve the problem.

Book Your Seat & Save 35%

Not Every Business Problem Needs AI

One refreshing takeaway from the session was that Prithvi repeatedly cautioned against applying AI indiscriminately.

Some business problems simply don’t require large language models.

Static reporting rarely does.

Rule-based automation often doesn’t.

Basic workflows can frequently be solved without introducing generative AI at all.

According to Prithvi, AI becomes valuable when decisions are repetitive, time-sensitive, involve multiple signals, or require synthesising complex information into actionable recommendations.

That’s where AI shifts from being a novelty to becoming genuine decision support.

Why Product Analytics Matters After Deployment

Perhaps the biggest insight from the webinar came after the AI product had already been deployed.

Shipping the product isn’t the finish line.

It’s the beginning.

Once users begin interacting with an AI assistant, product teams need to understand how that interaction unfolds.

Did users open the recommendation?

Did they read the explanation?

Did they accept the suggested action?

Did they complete it?

Where did they abandon the workflow?

These behavioural questions are often impossible to answer using traditional business intelligence alone.

Instead, they require product analytics.

Prithvi introduced tools such as Amplitude as examples of behavioural analytics platforms that capture user interactions throughout the product lifecycle. Unlike BI dashboards that explain what happened in the business, product analytics helps explain how users behave inside the product itself—and where trust begins to break down.

Where SQL and Python Fit Into AI Product Analytics

While AI dominated the discussion, the webinar repeatedly returned to two technologies that every data professional already knows.

SQL.

Python.

SQL forms the backbone of product analytics by transforming raw event logs into meaningful user funnels.

How many recommendations were generated?

How many were viewed?

How many were accepted?

How many ultimately resulted in completed actions?

Those simple funnel analyses often reveal where products lose users long before sophisticated machine learning becomes necessary.

Python then extends those insights through cohort analysis, segmentation, experimentation, A/B testing, behavioural modelling, and recommendation analysis.

Rather than replacing traditional analytics, AI product development depends on it.

As Prithvi put it, SQL helps measure the funnel, while Python helps explain the patterns behind it.

Measuring What Actually Matters

The webinar concluded with a useful framework for thinking about AI success.

At the bottom sit model metrics.

Precision.

Recall.

Latency.

Cost.

Groundedness.

These remain important.

Above them sits product adoption.

Are recommendations viewed?

Accepted?

Executed?

Are users providing positive feedback?

At the very top sit the metrics every business ultimately cares about.

Did stock-outs decrease?

Was waste reduced?

Were analysts able to make decisions faster?

Did customer experience improve?

Those are the outcomes that determine whether an AI product has truly succeeded.

Final Thoughts

If there was one message that carried through every section of the webinar, it was this:

Great AI products don’t simply generate better answers. They reduce decision friction.

As AI becomes deeply integrated into enterprise products, data professionals will increasingly find themselves working across product management, analytics engineering, experimentation, behavioural analytics, SQL, Python, and machine learning.

The organizations that succeed won’t necessarily be those with the most advanced models.

They’ll be the ones that understand their users best.

Because in the end, successful AI isn’t measured by what the model produces.

It’s measured by the decisions people make because of it—and the business outcomes those decisions create.

See you next time!

This NINJIO Insights Report dives into the key emotional susceptibilities that make social engineering work and offers concrete steps that your security team can take to equip your workforce to resist cyberattacks.

Download the Guide

👋 Hi there, welcome to DataPro #176.

As foundation models become increasingly commoditized, competitive advantage is shifting somewhere less glamorous but far more consequential: the quality of your data, features, and production systems.

This week’s lead story explores exactly that. In an insightful conversation, Cornellius Yudha Wijaya, Chief Product Officer and AI practitioner, shares why the features that deliver the best offline metrics often become the first ones to fail in production. Interviewed by Vaideeshwari Roshan, he explains why point-in-time correctness, feature stability, and engineering discipline matter far more than squeezing out another percentage point of model accuracy.

This week’s highlights:

• 🎙️ Expert Interview: Why production-ready feature engineering is about engineering trust, not just better features

• ☁️ Google Cloud’s Gemini Enterprise Agent Platform brings secure MCP connectivity to external AI agents and developer workflows

• 🐍 Baidu’s CUP toolkit streamlines production Python with logging, concurrency, caching, and reliability utilities

• 🧬 Schrödinger and Google DeepMind achieve a 4× acceleration in molecular simulations using AlphaEvolve

• 🤖 Claude Sonnet 5 pushes agentic coding forward with stronger benchmarks and compelling cost-performance gains

• 🛡️ Amazon Bedrock introduces centralized model governance and resilient LLM inference architectures for production AI

• 🎬 Outpost VFX reduces AI model training from weeks to days using distributed multi-GPU training on AWS

If you’re building machine learning systems that need to survive production, scale across cloud infrastructure, or keep pace with the rapidly evolving AI ecosystem, this edition is packed with ideas, architectures, and lessons worth borrowing.

Let’s get into it.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Book Your Spot & Save 35%

The Feature That Looked Perfect… Until It Reached Production

Feature engineering is often described as the secret sauce behind great machine learning models. But after reading through a conversation between my colleague Vaideeshwari Roshan and data scientist Cornellius Yudha Wijaya, I realized the conversation isn’t really about feature engineering. It’s about engineering trust.

Cornell has spent more than seven years building machine learning systems across insurance and AI startups, working with sensitive enterprise data where every prediction has business consequences. His perspective is refreshingly different from the feature engineering tutorials most of us consume. Rather than focusing on creating more features, he focuses on building features that survive production.

Below is an edited version of that conversation, along with a few observations that stood out to me as I read through it.

🎙️ Behind the Conversation

Interviewer: Vaideeshwari Roshan, Portfolio Manager, Packt

Expert: Cornellius Yudha Wijaya, Chief Product Officer, Data Scientist, and Author

“Feature engineering isn’t about creating more features.”

One of the first things that struck me was how quickly Cornell shifted the conversation away from algorithms.

Many of us instinctively think about machine learning in terms of model selection. Should we use XGBoost? CatBoost? Neural networks?

Cornell’s answer reminded me that production ML teams often spend far more time debating the quality of their features than the sophistication of their models.

Vaideeshwari: You’ve worked on enterprise machine learning systems as well as AI products in startups. How has your perspective on feature engineering evolved?

Cornellius: Early in my career, I thought feature engineering was mostly about creating predictive variables from raw data.

Over time, I realized that’s only one part of the job.

Today, I think about whether a feature is reliable enough for production. Can it survive upstream data changes? Is it explainable to business stakeholders? Can another engineer understand and reproduce it? Those questions matter just as much as predictive performance.

Feature engineering has become less about data manipulation and more about building reliable systems.

✍️ Editor’s Reflection

That distinction feels increasingly relevant.

As foundation models and AutoML continue to simplify model building, the competitive advantage is shifting elsewhere.

The harder problem isn’t choosing an algorithm.

It’s deciding which information your model should trust.

That philosophy became even clearer when Cornell described one of his largest production projects.

Building a churn prediction model where time was the biggest challenge

The project Cornell chose to discuss wasn’t unusual on paper.

It was a customer churn prediction system built at Allianz Life Indonesia.

The challenge wasn’t predicting churn.

The challenge was making sure every prediction reflected what the business actually knew at that moment in time.

Vaideeshwari: What made this project particularly challenging from a feature engineering perspective?

Cornellius: The data came from multiple enterprise systems spanning different time periods.

Every customer had to be represented at a specific reference date, and every feature had to be calculated only from information available before that point.

That sounds simple, but it’s one of the easiest places to make mistakes.

One incorrect join or timestamp can accidentally introduce future information into your feature set.

Vaideeshwari: You often emphasize “point-in-time correctness.” Why is it so important?

Cornellius: Because models should never learn from the future.

If a feature accidentally contains information generated after your prediction date, validation results become misleading.

The model appears much better than it really is.

Then you deploy it, and suddenly performance drops because that future information no longer exists.

Every feature should answer one question:

“Would we have known this information at prediction time?”

If the answer is no, the feature shouldn’t exist.

Book Your Seat & Save 35%

✍️ What surprised me most

Data leakage is something every ML engineer has heard about.

But Cornell frames it differently.

He doesn’t describe it as a modeling mistake.

He describes it as a feature engineering mistake.

That’s an important mindset shift because it moves the discussion upstream, where these problems actually begin.

The most valuable features weren’t static

One misconception I had before reading this interview was assuming customer profiles would carry most of the predictive power.

Instead, Cornell kept coming back to one word:

Behavior.

Vaideeshwari: Which features ultimately made the biggest difference?

Cornellius: Behavioral features consistently outperformed static customer attributes.

Instead of describing who customers were, we focused on how they were changing.

Rolling-window aggregates over the previous 7, 30, and 90 days helped capture recent activity.

Trend features showed whether engagement was increasing or decreasing.

Delta features highlighted meaningful behavioral shifts.

For highly skewed financial variables, log transformations also improved robustness.

Customers are constantly changing.

Our features needed to reflect that.

✍️ Editor’s Reflection

That answer stayed with me.

Good features don’t simply describe reality.

They describe change.

Whether you’re predicting churn, fraud, demand, or equipment failure, the strongest signal often isn’t the current state.

It’s the direction of movement.

The feature that improved accuracy… and still got deleted

This was probably the most unexpected moment in the conversation.

Cornell explained that his team deliberately removed features that improved model performance.

Naturally, Vaideeshwari asked why.

Vaideeshwari: Why would you remove a feature that improves your validation metrics?

Cornellius: Because validation metrics aren’t the business objective.

Some features looked extremely predictive because they contained subtle leakage.

Others depended on business rules that changed frequently.

A few simply weren’t stable over time.

Keeping them improved offline accuracy.

Removing them improved production reliability.

Sometimes you have to trade a slightly better benchmark for a much better production system.

✍️ One sentence worth remembering

A feature that boosts validation scores but can’t survive production isn’t an asset. It’s technical debt.

Production doesn’t end with model deployment

One theme kept appearing throughout the interview.

Feature engineering isn’t finished once the model is trained.

In many ways, that’s when the real work begins.

Vaideeshwari: How did your team maintain feature quality after deployment?

Cornellius: Every feature was treated like a managed asset.

We documented feature definitions, tracked version changes, maintained validation rules, and monitored data freshness, missing values, and business-rule consistency.

When upstream systems changed unexpectedly, anomaly detection alerted us before the issue affected production predictions.

Feature engineering is an ongoing operational process.

Vaideeshwari: Did richer features ever conflict with operational requirements?

Cornellius: Definitely.

More sophisticated features often improve predictive performance.

But they also increase computation costs, latency, and pipeline complexity.

We had to balance model quality with operational SLAs and business value.

The best feature isn’t always the most sophisticated one.

It’s the one that delivers reliable value at scale.

Three lessons every ML engineer can borrow

Before wrapping up, Vaideeshwari asked Cornell what he hoped practitioners would remember most from his experience.

Vaideeshwari: If our readers remember only three lessons from this conversation, what should they be?

Cornellius:

1. Point-in-time correctness is non-negotiable.

Never allow your model to learn from information it wouldn’t have during prediction.

2. Stability is a feature.

Reliable, consistent features often outperform clever ones over the long term.

3. Every feature should support a business decision.

Predictions become valuable only when they enable action.

Continue reading the full conversation on Packt’s Medium publication.

Book Your Seat & Save 35%

Data Science & ML Research Roundup

◾ Gemini Enterprise Agent Platform remote MCP server: Google Cloud has introduced the Gemini Enterprise Agent Platform remote MCP server, giving developers a secure way to connect external AI agents and IDEs like Claude Code to Google Cloud resources. Built on the open MCP standard, it enables access to Model Garden, prompt libraries, notebooks, and model management through a single governed interface, combining faster development with enterprise-grade security and IAM-based access controls.

◾ CUP (Common Useful Python): Building Reliable Python Workflows with Baidu’s Utility Toolkit. Baidu’s Common Useful Python (CUP) toolkit brings together production-ready utilities that simplify reliable Python development. From structured logging, configuration management, caching, and concurrency to ID generation, scheduling, networking, and resource monitoring, CUP provides a unified toolkit for building maintainable, scalable applications. The tutorial walks through practical workflows that help developers reduce boilerplate and improve the reliability of real-world Python systems.

◾ How Schrödinger sped up molecular discovery by 4x with Alphaevolve: Schrödinger has accelerated molecular simulations by 4× using Google DeepMind’s AlphaEvolve, an evolutionary AI coding agent that optimizes performance-critical algorithms in machine-learned force fields (MLFFs). By replacing computational bottlenecks with AI-generated parallel implementations, the company significantly sped up model training and inference, enabling faster drug discovery, catalyst design, and materials research while paving the way for AI-optimized GPU kernels in scientific computing.

◾ Anthropic Claude Sonnet 5 vs Sonnet 4.6 vs Opus 4.8: Agentic Coding Benchmarks, API Pricing, and Cost-Performance Tradeoffs Compared. Anthropic has unveiled Claude Sonnet 5, its most capable mid-tier model yet, designed for long-running agentic workflows, autonomous coding, and tool use. It outperforms Sonnet 4.6 across every published benchmark while narrowing the gap with Opus 4.8. With introductory API pricing of $2/$10 per million tokens, Sonnet 5 offers one of the strongest cost-to-performance ratios for AI coding, automation, and enterprise development workloads.

◾ Simplify multi-account access to Amazon Bedrock models with managed entitlements: Amazon Bedrock now supports managed entitlements, allowing enterprises to subscribe to third-party foundation models such as Anthropic Claude and Cohere once and securely distribute access across AWS accounts. By centralizing subscriptions with AWS License Manager, organizations simplify governance, streamline multi-account AI deployments, maintain consistent pricing, and eliminate the need for AWS Marketplace permissions in individual workload accounts.

◾ Implementing resilience patterns with Amazon Bedrock and LLM gateway: As generative AI moves into production, Amazon Bedrock provides a set of resilience patterns to keep LLM inference highly available, scalable, and cost efficient. The guide covers five production-ready approaches, including cross-Region inference, multi-account quota isolation, and intelligent request routing, helping organizations improve availability, handle traffic spikes, reduce throttling, and build reliable multi-model AI applications on AWS.

◾ How Outpost VFX Uses AWS to Accelerate AI Model Training for Visual Effects: Outpost VFX has accelerated AI-powered face replacement training by up to 8× using AWS multi-GPU P5 instances and PyTorch Distributed Data Parallel. By replacing single-GPU workflows with distributed training, the studio reduced model training from 1–2 weeks to just two days, enabling faster creative iteration, higher-quality outputs, and more efficient visual effects production at scale.

See you next time!

Build self-service analytics 4× faster with OpenUI Cloud.

What if every user could turn a question into a dashboard? OpenUI Cloud plugs into your existing data stack, transforming plain-English queries into live charts, KPI cards, reports, and dashboards—without adding to your BI backlog or frontend roadmap.

Start Building Now

👋Hey there, welcome to DataPro #175.

Building AI systems is no longer the hard part. Building AI systems that behave consistently, scale across teams, and are fast enough for production is. Many organizations are discovering that great prompts alone don’t translate into reliable products. What they need are reusable workflows, measurable infrastructure, and architectures designed for production from day one.

That’s exactly where this week’s leading story comes in. Elisa Terumi, PhD, explains why Skills are emerging as a foundational building block for AI agents, showing how they transform one-off prompts into reusable, version-controlled capabilities that make agentic systems easier to build, maintain, and scale.

In this week’s highlights:

• 🧠 Learn how AI Skills are reshaping agent architectures and why they’re becoming the new standard for reusable AI workflows.

• ⚡ Explore DFlash, a new speculative decoding technique delivering up to 15× higher LLM throughput on NVIDIA Blackwell GPUs.

• 📄 Discover Datalab’s lift and Baidu’s Unlimited OCR, two approaches pushing document AI forward with structured extraction and long-document parsing.

• 🎙️ See how speech AI is evolving with Gradium’s real-time translation models and Amazon Nova 2 Sonic powering natural voice agents.

• Discover how OpenUI Cloud brings conversational analytics to your existing data stack.

• 📊 Dive into the latest cloud tooling, from BigQuery Managed Python UDFs and Google Observability Analytics to enterprise-scale AI deployments across banking and telecom.

Whether you’re building agentic applications, optimizing inference, or deploying AI into production, this edition brings together the architectural patterns, infrastructure updates, and engineering breakthroughs shaping the next generation of Data and AI systems.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

⚙️ This Week's Packt Expert Workshop: Production-Ready GenAI Starts with Evaluation

As GenAI becomes part of AI applications, data pipelines, and ML workflows, evaluating model outputs is no longer optional. Join Amy Chen and Surjeet Mishra to learn practical frameworks for measuring LLM quality, identifying failure modes, building evaluation pipelines, and implementing feedback loops that make GenAI systems more reliable, trustworthy, and production-ready.

🎟 Save 30% today with the registration link.

The workshop is filling up quickly. If GenAI is part of your AI, ML, or data stack, you won't want to miss it.

Join Amy Chen & Sujeet Mishra Live

What Are Skills in AI Agent Systems? And How to Build Your Own

Written by Elisa Terumi, PhD

The term sounds simple. But in modern AI systems, it has a very specific meaning.

And understanding it changes how you build with LLMs.

What are skills?

Skills are modular, reusable instruction sets that teach an AI system how to perform specific tasks or workflows.

In systems like Claude Code, a skill is typically:

A folder

Containing a SKILL.md file

With structured instructions describing how a task should be executed

Once defined, the system can automatically apply that knowledge whenever a relevant request appears.

This is the key shift:

Instead of repeating prompts, you encode behavior once — and reuse it.

I’ve created a repository with practical examples of skills — feel free to explore it here: https://github.com/elisaterumi-ai/agent-skills-in-practice

Skills as structured capabilities (not just prompts)

A common misconception is to treat skills as “saved prompts.”

They are not. A saved prompt is a one-off instruction you reuse manually. A skill is closer to a standard operating procedure (SOP) for AI:

It defines what to do

When to do it

How to do it consistently

The practical difference is significant. A prompt depends on you remembering to use it and applying it correctly each time. A Skill is activated automatically by the system when the context is relevant, follows a testable structure, and can be shared with your team as part of the repository.

Technically, a Skill combines instructions, workflows, and context to handle multi-step tasks — while a standalone tool executes one specific deterministic function, and a one-off prompt has no structure or reuse.

That combination is what makes Skills an architectural pattern, not just a convenience.

How skills work under the hood

The execution model is subtle — and important.

When a system (like Claude Code) runs:

It loads only skill names and descriptions

It receives a user request

It performs semantic matching

It selects relevant skills

It loads the full instructions and executes them

This has two implications:

Skills do not clutter the context window

They activate only when needed

Skills vs prompts vs tools

Understanding this distinction is critical.

Prompts

One-off instructions

Not reusable

No structure

Tools

Execute a specific function

Deterministic behavior

Skills

Combine: instructions + workflows + context

Handle multi-step tasks

In other words:

A tool does one thing. A skill orchestrates how things should be done.

Why skills matter (from experimentation to production)

Skills are not just a convenience feature. They are an architectural pattern.

They enable:

Consistency → same output format every time

Reuse → define once, apply everywhere

Scalability → move from prompts to systems

Collaboration → share workflows across teams

In fact, skills are increasingly used to:

encode coding standards

enforce documentation formats

automate workflows

embed domain knowledge into AI systems

Where skills live

Skills are typically scoped at two levels:

Personal skills

Stored locally

Reused across projects

In Claude systems, personal skills live in ~/.claude/skills in your home directory. These follow you across all your projects — your commit style, your documentation format, how you like code explained.

Project skills

Stored in repositories

Version-controlled

Shared with teams

Project skills live in .claude/skills inside the repository root. Anyone who clones the repo gets these skills automatically. This is where team standards live: coding conventions, brand guidelines, project-specific processes. Because they sit inside the repository, they’re version-controlled alongside the code and shared naturally through Git.

This makes them part of the codebase — not just user configuration.

The Anatomy of a Skill

A Skill is a directory containing a SKILL.md file. The directory name should match the skill name. The file has two parts: a YAML metadata block at the top and Markdown instructions below.

The metadata defines name and description,both required. The description is the most critical field: it’s what Claude uses to decide whether the Skill is relevant. Two optional fields also exist: allowed-tools, which restricts which tools Claude can use while the Skill is active, and model, which specifies which Claude model to use for that Skill.

The instructions define the steps, rules, and output format. This is where the actual procedure lives.

When should you create a skill?

A practical rule:

If you are repeating the same instructions more than once, you should create a skill.

Typical use cases:

Code review guidelines

Commit message formats

Documentation templates

Data processing pipelines

Domain-specific transformations

Practical Example: A PR Description Skill

Let’s build a personal Skill that teaches Claude to write pull request descriptions in a consistent format.

First, create the directory:

mkdir -p ~/.claude/skills/pr-description

Then create the SKILL.md file inside that directory:

---

name: pr-description

description: Writes pull request descriptions. Use when creating a PR,

writing a PR, or when the user asks to summarize changes for a pull request.

---
When writing a PR description:

1. Run `git diff main...HEAD` to see all changes on this branch

2. Write a description following this format:

## What
One sentence explaining what this PR does.

## Why
Brief context on why this change is needed.

## Changes
- Bullet points of specific changes made
- Group related changes together
- Mention any files deleted or renamed

Restart Claude Code.

The next time you say “write a PR description for my changes,” Claude will recognize the request, load the Skill, and follow the template — same format every time.

Dive deeper into the topic on Packt’s Medium handle.

🕸️ Turn Connected Data Into Better AI Answers

Many RAG systems fail not because the model lacks knowledge, but because retrieval lacks structure. Join Bruno Gonçalves for a practical workshop on GraphRAG and learn how to build AI applications that can reason across relationships, answer multi-hop questions, and generate more trustworthy responses.

🎟️ Save 35% on your ticket with the DataPro community offer.

Discover how leading teams are tackling AI hallucinations and building systems that can reason across complex business data.

Save Your Spot Now!

Data Science & ML Research Roundup

◾ How Loka Built a Natural, Low-Latency Voice Agent with Amazon Nova 2 Sonic: Loka built a conversational AI agent with Amazon Nova 2 Sonic to eliminate the slow, robotic experience of traditional voice assistants. By using native speech-to-speech processing, the solution delivers faster responses, higher speech reasoning accuracy, and lower costs. Prompt engineering further boosted conversational quality, making the AI more natural, accurate, and production-ready for customer support at scale.

◾ Baidu Releases Unlimited OCR, a 3B Model That Keeps the KV Cache Flat for Long-Document Parsing: Baidu has open-sourced Unlimited OCR, a 3B-parameter model that solves a major OCR bottleneck by keeping memory usage constant, enabling efficient parsing of long documents in a single pass. Built on DeepSeek OCR, it delivers higher accuracy, faster throughput, and lower latency, making it well suited for large-scale document processing, transcription, and multimodal parsing workflows.

◾ Huntington Bank: Redacting sensitive data from 400M+ documents with AWS Huntington Bank cut a multi-year compliance project down to months by building a scalable AWS-powered pipeline to detect and redact sensitive data across 400 million documents. Using Amazon Textract, SageMaker, Step Functions, and Lambda, the solution achieved over 95% redaction accuracy while securely processing documents at massive scale with high concurrency and PCI DSS compliance.

◾ Datalab Releases lift: A 9B Open-Weights Vision Model That Extracts Structured JSON From PDFs Using Schemas Datalab has introduced lift, a 9B open-weights vision model that extracts structured JSON from PDFs and images using JSON schemas. It achieves 90.2% field accuracy while processing multi-page documents in a single pass, making it one of the strongest self-hostable extraction models. Schema-constrained decoding reduces hallucinations and enables reliable document automation workflows.

◾ Build a healthcare appointment agent with Amazon Nova 2 Sonic AWS has published a reference architecture for building healthcare appointment agents with Amazon Nova 2 Sonic and Bedrock AgentCore. The speech-to-speech AI authenticates patients, confirms or reschedules appointments, collects pre-visit information, and escalates to staff when needed. Built with serverless AWS services and healthcare-specific tools, it enables natural, low-latency voice interactions that can help reduce appointment no-shows at scale.

◾ Gradium Launches stt-translate and s2s-translate, Real-Time Speech Translation Models Beating gpt-realtime-translate on Accuracy and Latency: Gradium has launched stt-translate and s2s-translate, real-time speech translation models that combine transcription, translation, and speech output into a faster two-model pipeline. Supporting five languages and 20 pairs, the models claim stronger BLEU accuracy than GPT and Gemini alternatives, 3-second average latency, live browser streaming, and voice control, including cloning for multilingual meetings, agents, and dubbing.

◾ Open models, global networks: How AT&T and GSMA are accelerating innovation with Gemma Google Cloud and GSMA have introduced Open Telco AI, an initiative built on Gemma models to bring domain-specific AI to telecom networks. Fine-tuned on specialized telecom data, the open OTel models outperform larger general-purpose models on network tasks while reducing hallucinations through RAG. The project aims to accelerate AI-driven network automation, self-healing systems, and telecom-grade AI adoption.

◾ How to Design an OpenHarness Style Agent Runtime with Tools, Memory, Permissions, Skills, and Multi-Agent Coordination: This tutorial breaks down how to build an OpenHarness-style agent runtime from scratch, exposing the full mechanics behind modern agent systems. It walks through tool schemas, permissions, lifecycle hooks, memory, skills, retries, cost tracking, context compaction, and multi-agent coordination, giving developers a runnable framework for understanding how agents reason, call tools, manage state, and complete tasks.

◾ Query logs and traces with SQL in Observability Analytics: Google Cloud has rebranded Log Analytics as Observability Analytics, adding GA support for SQL-based analysis of logs and traces in a unified workspace. Developers can now join telemetry with business data to troubleshoot applications, optimize AI agents, and identify performance bottlenecks using BigQuery-powered SQL, while the new Observability API enables programmatic access for agentic workflows and automation.

◾ DFlash Speculative Decoding Drafts Whole Token Blocks in Parallel for Up to 15x Higher Throughput on NVIDIA Blackwell: Researchers at UC San Diego have introduced DFlash, a speculative decoding method that generates entire token blocks in parallel instead of one token at a time. By combining lightweight diffusion drafting with autoregressive verification, DFlash delivers up to 6× faster lossless inference in research benchmarks, while NVIDIA reports up to 15× higher throughput on Blackwell GPUs for latency-sensitive AI workloads such as coding agents and reasoning models.

◾ Python UDF in BigQuery, now generally available: Google Cloud has announced the general availability of BigQuery Managed Python UDFs, enabling developers to run custom Python code and popular libraries like NumPy, pandas, and scikit-learn directly within BigQuery SQL. The serverless feature eliminates infrastructure management while supporting vectorized execution, configurable compute resources, external API integration, and production-grade monitoring for advanced analytics and machine learning workflows.

See you next time!

Welcome to DataPro #174.

This week, we’re exploring a question that sits at the heart of enterprise AI: Why do RAG systems still hallucinate when the answer is already in the data? Former JPMorgan Chase VP of Data Science Bruno Gonçalves unpacks the retrieval challenges behind many RAG failures and explains why GraphRAG is gaining traction for multi-hop reasoning, corpus-wide analysis, and explainable AI.

If you're building AI applications today, these challenges will likely feel familiar. From unreliable retrieval and reasoning gaps to measuring AI ROI, scaling agents, and managing ever-growing context windows, organizations are moving beyond experimenting with AI and into the harder work of making it reliable, explainable, and production-ready. Many of the stories in this week's edition explore that transition from prototype to practice.

As part of our commitment to bringing practical, practitioner-led learning to the DataPro community, we’re also pleased to partner with GrowthSchool and Outskill. Together with AI communities around the world, we're helping surface high-quality learning opportunities, from free modules and workshops to expert-led training and hands-on sessions hosted by the Packt Virtual Conference team.

One such opportunity this week is a free workshop by Outskill on turning Claude into a 24/7 AI copilot, covering Skills, Connectors, Cowork, vibe coding, and workflow automation. It’s a practical, hands-on way to learn how experienced practitioners are integrating AI into their daily workflows. And if you can’t join live, every session is recorded and shared with registered attendees, so you can learn at your own pace and revisit the material whenever you need it.

Make Claude Your 2nd Brain That works 24/7 By Mastering In It 16 hours

Register Now!

This Week’s Highlights

🔬 OpenAI’s LifeSciBench Reveals How Far AI Still Is From Scientific Reasoning
A new 750-task benchmark built by 173 PhD scientists shows even the best models pass just 36% of real-world life sciences research tasks.

🤖 Vercel Open-Sources Eve, the Agent Framework Behind 100+ Production AI Agents
A filesystem-first framework where agents are simply directories of tools, skills, channels, schedules, and subagents.

📈 Google’s DORA Research Says AI ROI Starts With a Productivity Dip
Why the “J-Curve” of AI adoption is normal, and how leading engineering teams are measuring the business impact of AI.

⚡ Amazon SageMaker Async Inference Removes Mandatory S3 Uploads
A new inline payload feature cuts latency, simplifies architecture, and reduces costs for asynchronous AI workloads.

🧠 MiniMax-M3 Launches With 1 Million Context and Sparse Attention Breakthroughs
The 428B-parameter multimodal model delivers 9× faster prefilling and 15× faster decoding at million-token scale.

🚀 GLM-5.2 Raises the Bar for Open-Source Agentic Coding Models
Featuring a stable 1M-token context window, improved coding benchmarks, and an MIT license.

💻 Run a Powerful Coding Model Locally With Just 4.5GB of VRAM

Gemma4-12B-Coder combines verified reasoning traces and coding expertise in a lightweight, fully local package.

Let’s get into it.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Why RAG Hallucinates With the Answer in Hand

Written by Bruno Gonçalves, data scientist, educator, and former Vice President of Data Science and Finance at JPMorgan Chase.

Vector RAG has a strange failure pattern. The answer sits in your documents; the retriever returns something plausible but still gets it wrong. The fault sits with your retrieval scaffolding that cannot see structure, and it shows up in three predictable ways.

1. Multi-hop questions. Ask about the indirect suppliers of Company X. The retriever returns chunks that mention Company X, but the indirect suppliers live in documents that never name it. One document says firm A supplies firm B. Another says firm B supplies firm C. No single chunk holds the chain, so no similarity score will surface it. Cosine similarity has no notion of A → B → C. A knowledge graph stores those links as data: ndes represent Entities and their relationships are edges. With this formalism, the question is just a two-hop traversal that runs in milliseconds.

2. Global questions. “What are the main themes across these 500 documents?” Top-k retrieval grabs the ten chunks closest to the query and ignores the other ten thousand. The GraphRAG paper named the problem: global questions are summarization tasks, not retrieval tasks. The fix is to build the a map first, then compute the answers from the map. Extract entities and relations from every chunk, cluster them into communities, summarize each community, then reduce those summaries into one response. On million-token test corpora, this beat vector RAG on both the comprehensiveness and the diversity of its answers.

3. Explainability. Vector RAG can show you a relevant chunk. It cannot show you a reason. A chunk might score 0.87 on query similarity, but that is the whole story. Graph RAG returns the chain of entities and relations behind the answer, hop by hop, back to the source text. In finance, healthcare, and law, that audit trail decides whether the system ships.

The thread running through all three failure modes is the same. Hallucination in RAG is rarely a generation problem. It is a retrieval problem wearing a generation costume. The model invents connections precisely where the retriever failed to supply them: across hops, across the whole corpus, and across the gap between an answer and its evidence. Hand the model real structure instead of a stack of similar-looking chunks, and the invented connections mostly disappear. The documents had the answer all along. The graph is what lets the system find it.

Full GraphRAG is expensive to index. The pipeline makes LLM calls for every chunk, every entity description, and every community summary. One practitioner account puts the price of indexing a single five-gigabyte legal dataset at $33,000 in early 2024. That number kept plenty of teams on plain vectors, whatever the quality argument said.

LazyGraphRAG attacked the cost in November 2024. It skips upfront summarization entirely, relying instead on noun-phrase extraction and co-occurrence statistics, with zero LLM calls, so it costs about the same as building a vector index, or about 0.1% of full GraphRAG. The graph work shifts to query time, where an iterative search builds only the structure a given question needs. Answer quality matches GraphRAG global search at more than 700 times lower query cost, and the cost objection to graphs largely dissolved.

Want to Build a GraphRAG System?

Reading about GraphRAG is one thing. Building one is another.

Start with [our walkthrough], where you’ll learn how to turn 2,000 news articles into a searchable knowledge graph. Then take the next step on July 11 with our live Production GraphRAG Workshop, where you’ll build a complete GraphRAG chatbot from raw Wikipedia data in just 3.5 hours.

You’ll learn entity and relationship extraction with spaCy and REBEL, graph construction with NetworkX, hybrid graph-plus-vector retrieval, and grounded generation with an LLM. You’ll also receive the recording, source code, slides, and a certificate, so you can revisit the material long after the session ends.

🎟️ Exclusive for DataPro readers: Save 35% on your ticket.

Whether you’re exploring GraphRAG for the first time or looking to improve an existing RAG pipeline, this workshop will give you a practical framework for tackling the multi-hop questions traditional retrieval systems often miss.

Basic Python and Docker knowledge are all you need. Seats are limited.

Book Your Spot & Save 35%

Data Science & ML Research Roundup

◾OpenAI Releases LifeSciBench, a 750-Task Benchmark Grading AI Models on Real Life-Science Research With Expert-Written Rubric: OpenAI has launched LifeSciBench, a new benchmark designed to test how well AI models perform real-world life sciences research, not just fact recall. Built by 173 PhD-level scientists, it includes 750 expert-authored tasks spanning 7 research workflows and 7 biological domains, supported by 1,062 artifacts and graded against 19,020 rubric criteria. Results show significant headroom: the top-performing model, GPT-Rosalind, achieved a 36.1% pass rate and 0.576 normalized score, while GPT-5.5 scored 25.7%. Models struggled most with artifact-heavy tasks, experimental design, and precise sequence generation, highlighting how far AI remains from matching expert scientific reasoning.

◾Vercel Releases Eve: An Open-Source AI Agent Framework Where Each Agent is a Directory of Files Mapped to Capabilities: Vercel has open-sourced eve, the AI agent framework powering 100+ production agents internally. Built around a filesystem-first approach, each agent is simply a directory where files map to capabilities like tools, skills, channels, schedules, and subagents. The framework ships with built-in durability, sandboxed execution, human approvals, secure MCP/OpenAPI connections, multi-channel deployment, and observability. Vercel reports real-world adoption at scale: d0 handles 30,000+ data queries monthly, Vertex autonomously resolves 92% of support tickets, and its autonomous SDR delivers a reported 32× ROI. Agents can be scaffolded with a single command and deployed unchanged from local development to production.

◾How to measure the business value of generative AI: Google’s latest DORA research on AI-assisted software development argues that proving AI ROI requires more than measuring productivity gains. The report highlights a “J-curve” effect, where teams often experience an initial productivity dip due to learning new workflows, increased code-review demands, and bottlenecks in testing and approvals. While 90% of surveyed developers now use AI at work, financial outcomes vary widely, with successful organizations investing in workflow and cultural changes alongside tooling. DORA recommends building explicit ROI models that account for both visible costs and hidden adoption challenges, using frameworks and calculators that link AI investments to productivity, security, developer experience, and business growth.

◾Amazon SageMaker AI Async Inference now supports inline request payloads: Amazon SageMaker AI has added inline payload support for Async Inference, allowing developers to send request data directly through the new Body parameter instead of first uploading inputs to Amazon S3. The feature supports payloads up to 128 KB, eliminates an extra network round-trip, removes S3 upload costs and IAM dependencies, and simplifies client code while maintaining existing output behavior through S3. Designed to be backward compatible, it works with existing async endpoints without model or container changes and is available across 31 AWS commercial regions. AWS recommends inline payloads for smaller JSON and structured-data workloads, while larger inputs such as images and audio should continue using S3-based InputLocation workflows.

◾MiniMaxAI/MiniMax-M3: MiniMax has released MiniMax-M3, a native multimodal foundation model with a massive 1 million-token context window, 428B total parameters, and 23B activated parameters. Unlike models that add multimodal capabilities later, M3 is trained on text, images, and video from the first training step, enabling deeper cross-modal reasoning. Its new MiniMax Sparse Attention (MSA) architecture dramatically improves long-context efficiency, delivering 9× faster prefilling, 15× faster decoding, and reducing per-token compute costs to 1/20th of its predecessor (M2) at 1M-token context. The model also targets advanced coding and agentic workflows, achieving frontier-level results on long-horizon coding and cowork benchmarks, and supports three reasoning modes—enabled, adaptive, and disabled—to balance accuracy, latency, and throughput. Since its release, the open-source model has already recorded 56,000+ monthly downloads on Hugging Face.

See you next time!

Welcome to DataPro #173 👋

As we embark on this week’s edition, we want to extend a special welcome to all our new readers joining us from the AI Skills Conference organized by Community Sprints. We’re excited to have you as part of the Packt DataPro community — one of the industry’s largest and fastest-growing Data Science, Machine Learning, and AI newsletters with 125K+ data professionals worldwide.

At DataPro, we go beyond headlines. Every week, we break down the real operational bottlenecks shaping modern AI systems, explore expert-led solutions, and highlight the most important advancements across Data Science, ML engineering, GenAI, MLOps, and enterprise AI infrastructure.

As part of the Packt ecosystem, you’ll also gain access to exclusive perks, expert sessions, community connections, research-driven insights, curated learning resources, and future goodies designed specifically for AI practitioners, data scientists, engineers, and decision-makers. So, stay tuned — there’s a lot ahead.

This week’s edition also features an insightful expert-led deep dive from AI Engineer Esther Guru, who explains why most ML systems fail in production and what it actually takes to build reliable, scalable machine learning infrastructure beyond model training. From data drift and monitoring to operational reliability and deployment trade-offs, the session delivers practical lessons every ML engineer should understand before shipping AI systems into the real world.

Alongside that, this week’s research and engineering roundup covers some of the most important developments shaping the future of AI infrastructure and agentic systems:

• Stability AI’s Stable Audio 3, introducing scalable latent diffusion for long-form audio generation

• Google’s Agentic Enterprise vision, featuring Gemini 3.5, Gemini Omni, Antigravity, and Gemini Spark

• OmniVoice Studio, an open-source local alternative to ElevenLabs for voice AI workflows

• Google Agent Executor, a new distributed runtime for reliable long-running AI agents

• NVIDIA’s Gated DeltaNet-2, improving long-context reasoning in linear attention architectures

• Google AI Edge Portal, enabling on-device LLM benchmarking across 120+ Android devices

• AWS Bedrock AgentCore, powering scalable multi-agent orchestration, observability, and agentic commerce infrastructure

From multimodal AI and distributed agents to edge inference, voice AI, and production ML systems, this edition captures the accelerating shift toward scalable, operational AI engineering.

Let’s dive in. 🚀

Before we get into this week’s developments, join Amy Chen and Sujeet Mishra this weekend for an intensive hands-on workshop on building reliable GenAI systems with AI evaluations, observability, testing, and production-grade workflows. Learn how leading AI teams debug hallucinations, evaluate prompts and agents, monitor LLM systems in production, and build scalable evaluation pipelines for real-world AI applications.

If you’re building RAG systems, copilots, AI agents, or enterprise GenAI products, this session will provide practical frameworks and workflows you can immediately apply to improve reliability, performance, and shipping confidence.

⚡ Few seats left — claim your 15% discount by registering through the link below (discount already enabled).

Register Now

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

This Week’s Sponsor

Grow your Mac app with Setapp

Get up to 30K unique impressions right after launch while Setapp handles distribution, billing, licensing, taxes, and customer support.

You build great software. Setapp helps you grow revenue and reach the right users faster.

Join Setapp

Why Most ML Projects Fail in the Real World

AI engineer Esther Guru explains what it really takes to build reliable ML systems beyond just training models.

Beyond the Model: Building Real World Machine Learning Systems | From Training to Deployment — YouTube

Machine learning has become one of the defining technologies of modern software. From recommendation engines and fraud detection systems to healthcare applications and financial forecasting, ML is now deeply embedded into the products people use every day.

But despite all the excitement around artificial intelligence, the reality of deploying machine learning systems is far more complicated than most organizations expect.

Many machine learning projects never make it into production. Others reach deployment but slowly fail over time because the real world changes faster than the systems behind them. In many cases, companies spend enormous amounts of money building sophisticated models only to discover that the actual problem was never about the model itself.

That was the focus of a recent Packt DataML Talk hosted by Abhishek Kaushik, where AI engineer and electrical engineer Esther Guru explored the realities of engineering machine learning systems for production.

Instead of focusing only on algorithms, Esther explained how real-world ML systems behave outside research environments, why production systems fail, and what engineers need to understand before deploying machine learning at scale.

The session offered an important reminder:

Machine learning in production is not just about building models. It is about building systems.

When Should You Use Machine Learning?

One of the biggest mistakes organizations make is assuming that every problem requires AI.

According to Esther, machine learning should only be used when a problem satisfies a few important conditions.

First, the system must actually involve learning. If a problem can already be solved through a fixed equation or a simple rule-based system, machine learning may only add unnecessary complexity.

For example:

Ohm’s Law already provides a direct mathematical relationship between voltage, resistance, and current. There is no need for a neural network to solve such a problem.

Similarly, if a sports website simply wants to display the top-performing football players based on statistics, a sorting query is enough. Building a machine learning model for that task would be excessive.

Machine learning becomes useful when relationships between variables are too complex for traditional methods.

Problems such as predicting customer churn, forecasting housing prices, detecting fraud, or predicting the outcome of sports tournaments involve large numbers of interacting variables. These are the kinds of problems where ML systems become valuable.

Esther described machine learning as:

The process of learning complex patterns from existing data and using those patterns to make predictions on unseen data.

That definition highlights the most important requirement for ML systems: the ability to generalize.

A successful model is not one that memorizes training data. It is one that performs reliably on new data it has never seen before.

Research ML vs Production ML

One of the most valuable parts of the session was Esther’s explanation of the difference between research machine learning and production machine learning.

Many engineers assume that if a model performs well during experimentation, it is ready for deployment.

In reality, research ML and production ML operate under completely different conditions.

In research environments, engineers typically work with clean datasets, static files, and controlled experiments. Most datasets are already prepared and structured. Training happens in predictable conditions.

Production environments are very different.

Real-world data is messy, incomplete, noisy, and constantly changing. Data pipelines may fail. User behavior shifts. External events change the environment the model operates in.

This difference becomes especially important when discussing objectives.

In research, the main goal is usually simple: build the most accurate model possible.

In production, however, multiple teams have competing priorities.

An ML engineer may want a highly sophisticated deep learning system with maximum accuracy. Meanwhile, the business team may care more about infrastructure costs, scalability, and speed.

Production systems also prioritize latency.

Users expect predictions instantly. If a recommendation engine or AI assistant takes too long to respond, users abandon the platform.

This means production systems must optimize not only for accuracy, but also for:

• speed

• scalability

• cost efficiency

• reliability

• infrastructure performance

This is why production engineering matters just as much as model design.

Why Most ML Projects Fail

One of the most striking insights from the Packt Talk was how frequently ML projects fail.

Esther referenced industry findings showing that a large percentage of machine learning initiatives never even reach production. Even among deployed systems, many degrade significantly within months.

There are several reasons for this.

Poor Problem Framing

Many organizations begin AI projects without clearly defining the actual business problem.

Teams often say things like:

• “We want to use AI.”

• “We need a machine learning strategy.”

• “Can we automate this?”

But vague business goals lead to vague ML objectives.

Without clear alignment between technical teams and business requirements, organizations end up building systems that never deliver meaningful value.

According to Esther, successful ML projects begin with proper problem framing.

Teams must understand:

• whether ML is even necessary

• what success actually looks like

• which business metrics matter most

• what constraints exist in production

Without this clarity, even technically strong models can fail.

Catch the complete story and technical insights on Packt’s Medium handle.

Data Science & ML Research Roundup

◾Stability AI Releases Stable Audio 3: A Family of Fast Latent Diffusion Models for Audio Generation and Editing: Stability AI has open-sourced Stable Audio 3, a new family of latent diffusion models for generating and editing stereo audio at 44.1 kHz. The release includes scalable models, variable-length generation, fast inference, inpainting, and a novel SAME autoencoder with 4096× compression. SA3 supports music and SFX creation, achieves strong benchmark scores, and enables long-form audio generation efficiently on consumer hardware.

◾Innovations from Google I/O 26 on Google Cloud: At Google I/O 2026, Google expanded its vision for the “Agentic Enterprise,” where AI agents move beyond chatbots to autonomously execute workflows across apps, data systems, and developer environments. The announcement introduced Gemini 3.5 Flash for advanced reasoning and coding, Gemini Omni for multimodal video generation, Antigravity for enterprise-scale agent orchestration, and Gemini Spark, a personal AI work agent integrated across Workspace and business tools. Together, these launches signal Google’s push toward AI-native productivity, software development, and enterprise automation.

◾Meet OmniVoice Studio: A Local, Open-Source Alternative to ElevenLabs. OmniVoice Studio is emerging as a compelling open-source alternative to ElevenLabs, offering fully local voice AI capabilities without subscriptions or cloud processing. The desktop app supports voice cloning, video dubbing, diarization, dictation, and multilingual TTS across 646 languages. Built with FastAPI, WhisperX, Demucs, and Pyannote, it also includes an MCP server for integrations with Claude, Cursor, and custom AI workflows.

◾Agent Executor, Google’s distributed Agent Runtime: Google has open-sourced Agent Executor, a runtime standard designed to make long-running AI agents more reliable, scalable, and production-ready. Built for the emerging “agentic enterprise,” it introduces durable execution, secure sandboxing, session consistency, trajectory branching, and distributed deployment support. Combined with Agent Substrate for Kubernetes-scale orchestration, Google is positioning Agent Executor as foundational infrastructure for running millions of enterprise AI agents across hybrid environments.

◾NVIDIA AI Releases Gated DeltaNet-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule. NVIDIA has introduced Gated DeltaNet-2, a new linear attention architecture designed to improve long-context memory handling in AI models. By separating memory “erase” and “write” operations into independent channel-wise gates, the model achieves stronger retrieval accuracy and reasoning performance than Mamba-2, KDA, and prior DeltaNet variants. Trained at 1.3B parameters on 100B tokens, it also maintains efficient linear-time scaling for long-running sequence processing.

◾Benchmark LLMs on-device with AI Edge Portal: Google has expanded AI Edge Portal with new tools for benchmarking and debugging on-device LLMs across more than 120 Android device types. The update helps developers optimize latency, memory usage, and inference performance for mobile AI workloads using metrics like decode speed and initialization time. Google also introduced Model Explorer, a visualization tool for analyzing model graphs, quantization issues, and hardware compatibility in edge AI deployments.

◾Vibe-coded AI Studio apps with Firestore, Firebase, Cloud SQL: Google is expanding AI Studio into a full-stack “vibe coding” platform, enabling developers to build and deploy AI-powered applications directly to Google Cloud without needing billing setup or infrastructure management. The update adds support for Cloud SQL, Firestore, Firebase Auth, and Workspace integrations, while AI agents can now automatically provision databases, generate schemas, configure authentication, and deploy apps through natural language prompts.

◾Technical deep dive: AgentCore payments and innovation in agentic commerce. Amazon has introduced Bedrock AgentCore Payments, a managed infrastructure layer that enables AI agents to autonomously make secure microtransactions for APIs, MCPs, and digital services. Designed for the emerging “agentic economy,” the platform handles payment orchestration, wallet security, stablecoin support, spending guardrails, and observability through a single API. The release positions AgentCore as foundational infrastructure for scalable, enterprise-grade autonomous AI commerce.

◾Build highly scalable serverless LangGraph multi-agent systems in AWS with Amazon Bedrock AgentCore: AWS has introduced a serverless multi-agent AI architecture that combines LangGraph orchestration with Amazon Bedrock AgentCore Memory and Observability. The framework uses AWS Lambda, Step Functions, and Bedrock to build scalable AI workflows with persistent memory, real-time telemetry, and parallel agent coordination. Demonstrated through a marketing campaign review system, the solution highlights how enterprises can operationalize reliable, observable, and production-ready multi-agent AI systems on AWS.

◾Build high-performance generative AI systems with Strands Agents, NVIDIA NIM, and Amazon Bedrock AgentCore. AWS and NVIDIA have introduced a production-ready multi-agent AI architecture that combines NVIDIA NIM for GPU-accelerated inference, Strands Agents for orchestration, and Amazon Bedrock AgentCore for memory and observability. Designed for scalable enterprise AI systems, the framework supports parallel agent execution, persistent context, real-time monitoring, and serverless deployment, enabling organizations to build reliable, high-performance generative AI workflows at scale.

See you next time!

Welcome to DataPro #172 — this week, we explore the AI skills, agents, and production systems shaping 2026, alongside insights from the upcoming free AI Skills Conf happening tomorrow.

As AI rapidly moves beyond prompting into agentic workflows, context engineering, multimodal systems, and operational AI infrastructure, the real advantage now belongs to professionals who can build and deploy AI systems, not just use them. That’s why this edition features an expert-led deep dive from AI engineer Hari Prasad Renganathan on the 10 essential AI skills professionals need in 2026, covering RAG, LLMOps, AI evaluation, autonomous agents, and more.

The same shift is driving massive interest in the AI Skills Conf, where 6,000+ AI professionals are already registered to hear leaders from Google DeepMind, AWS, Meta, Spotify, SAP, DoorDash, and Scale AI discuss topics like “How to Become Irreplaceable with AI,” “Building Your AI Chief of Staff,” “Vibe Coding,” and “The Context Engineering and Agentic Memory.”

This is part of Packt DataPro’s Knowledge Partner initiative with global AI communities, complementing the hands-on workshops and technical deep dives delivered by the Packt Virtual Conference team. I’ll also be joining the “AI ROI Reality Check” panel alongside leaders from Spotify and SAP to discuss where enterprises are seeing measurable business value from AI today.

🚀 Registration is free, and even if you miss the live sessions, all recordings, AI workflows, templates, guides, and bonus resources will be delivered directly to your inbox afterward.

Register free and secure your spot

In this week’s highlights:

• OpenAI launched Daybreak for AI-driven cyber defense workflows
• BASF is using evolutionary AI to optimize global supply chains
• Sakana AI and NVIDIA introduced sparse transformer architectures for faster LLMs
• Google Cloud unveiled Cloud Storage Rapid for AI-scale infrastructure
• AWS and Exa demonstrated autonomous web research agents powered by real-time retrieval

The AI era is no longer just about using tools. It is about understanding the systems behind them.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

“Most Developers Are Learning AI the Wrong Way”: Hari Prasad Renganathan’s 10 Essential AI Skills for 2026

From prompt engineering and AI agents to LLMOps and multimodal systems, here’s the practical roadmap developers and AI professionals need to stay ahead in 2026.

Artificial Intelligence is no longer a futuristic concept reserved for research labs or billion-dollar tech companies. It is now woven into the everyday workflows of developers, startups, enterprises, product teams, and even non-technical professionals. In 2026, the biggest challenge is not whether AI will transform industries. It already has. The real question is this:

What should developers and AI professionals learn next to stay relevant?

The pace of AI innovation is moving faster than most people can keep up with. New frameworks emerge every month. AI models become more capable every quarter. Entire workflows are being automated in ways that seemed impossible just a few years ago. Yet amid all the noise, there are a handful of foundational topics that truly matter.

In a recent webinar hosted by Packt Talks, AI engineer and founder Hari Prasad Renganathan shared a practical roadmap for developers looking to thrive in the AI era. With experience spanning Columbia University, AI leadership roles at YC-backed startups, and building real-world AI products, Hari distilled the overwhelming AI landscape into ten critical topics every professional should understand.

Hosted by our Growth Lead, Abhishek Kaushik, the session focused not on hype, but on practical, production-ready AI skills that companies are already adopting today.

This article expands on those ideas into a complete guide for developers, engineers, and AI professionals preparing for the next wave of intelligent systems.

1. Advanced Prompt Engineering

Every AI workflow starts with prompts.

While basic prompting is now common knowledge, advanced prompt engineering has become a serious professional skill. Developers who understand how to structure prompts effectively can dramatically improve the quality, reliability, and relevance of AI outputs.

Hari emphasized an important point during the session:

“The best prompt you can give is better context.”

That single idea captures the essence of modern prompt engineering.

Today’s professionals must go beyond simple instructions and learn how to provide context, examples, reasoning patterns, and clear output expectations. Techniques such as few-shot prompting, chain-of-thought prompting, and structured outputs have become critical in enterprise AI systems.

Structured outputs are especially important because they allow AI systems to return clean JSON, XML, or CSV responses that integrate directly into applications and workflows. Prompt engineering may look simple on the surface, but in production systems, it directly impacts performance, cost, and user trust.

2. Retrieval-Augmented Generation (RAG)

Large Language Models are powerful, but they have limitations.

One major limitation is that they cannot naturally access an organization’s private knowledge base. That is where Retrieval-Augmented Generation, or RAG, becomes essential.

RAG allows AI systems to retrieve relevant information from documents, databases, PDFs, internal systems, or even web sources before generating a response. Instead of relying purely on pretrained knowledge, the model becomes context-aware and personalized.

Hari described RAG as one of the most practical skills AI engineers should learn because it bridges the gap between generic AI and real-world enterprise applications.

At its core, RAG works by converting information into vector embeddings, storing them in databases, retrieving the most relevant chunks during a query, and feeding those results back into the model. This creates AI systems that feel dramatically smarter and more useful.

Most enterprise AI tools today rely heavily on some form of retrieval system because businesses care less about generic intelligence and more about contextual accuracy.

3. Agentic AI Systems

If prompt engineering is the first level of AI maturity, agentic AI is the next major leap.

Traditional AI systems behave like assistants that answer questions. Agentic systems behave more like autonomous workers capable of completing tasks on behalf of users.

Hari described this transformation perfectly during the webinar:

“What if we can give hands and legs to that AI?”

That is the essence of AI agents.

Instead of simply generating text, agents can send emails, interact with APIs, manage workflows, run code, query databases, and use external tools. Developers can now create systems where AI decides what steps should happen next rather than following rigid workflows.

Modern frameworks such as LangGraph, CrewAI, and AutoGen are enabling developers to build increasingly sophisticated agentic systems. Hari specifically highlighted LangGraph as his preferred framework for production-grade AI workflows.

However, he also offered a realistic perspective. Despite the hype surrounding AI agents, most businesses are still primarily using prompt engineering and RAG because those systems are easier to deploy and maintain. Agentic AI is powerful, but the industry is still discovering its most commercially viable use cases.

4. Fine-Tuning Large Language Models

A few years ago, fine-tuning was considered one of the most important areas in AI development. Today, it has become more specialized.

Modern foundation models are already extremely capable. Many business problems can now be solved using better prompts, retrieval systems, and workflow orchestration without retraining models.

Still, fine-tuning remains valuable for highly customized applications.

Fine-tuning involves taking an existing model and retraining it on domain-specific data so it behaves in a more tailored way. Industries such as finance, healthcare, and legal services often benefit from fine-tuned systems because they require strict consistency and domain-specific behavior.

Hari pointed out that fine-tuning introduces operational complexity. Beyond training costs, organizations must also handle hosting, scaling, monitoring, and infrastructure management. That is why many companies prefer retrieval-based systems before committing to full fine-tuning pipelines.

Even so, developers should still understand fine-tuning because it remains an important tool for advanced AI applications. Platforms like Hugging Face have made experimentation with open-source models significantly more accessible.

5. LLMOps and AI Production Systems

Building an AI demo is easy. Deploying AI reliably at scale is the real challenge.

That challenge is where LLMOps comes in.

LLMOps focuses on operationalizing AI systems in production environments. Developers must think beyond experimentation and address real-world engineering concerns such as scalability, latency, reliability, monitoring, and governance.

Hari highlighted a key difference between traditional software and AI systems. Traditional systems are deterministic, meaning the same input produces the same output every time. AI systems behave differently. Even the same prompt can produce slightly different responses at different times.

This creates entirely new operational challenges.

Production-grade AI systems therefore require observability tools, prompt versioning systems, human review workflows, and robust monitoring pipelines. Tools like LangSmith and LangFuse are increasingly becoming essential for debugging and tracking AI behavior in production.

Without proper LLMOps practices, even impressive AI products quickly become unstable, expensive, and difficult to maintain.

6. AI Evaluation Frameworks

One of the hardest problems in AI engineering is evaluation.

How do you determine whether an AI system is actually performing well?

Traditional software testing relies on predictable outputs, but AI systems are probabilistic by nature. This makes evaluation significantly more difficult.

Hari explained that a major portion of his professional work involved building evaluation frameworks for AI systems. These frameworks help teams measure output quality, detect hallucinations, monitor performance degradation, and maintain alignment with business goals.

One popular approach is “LLM-as-a-Judge,” where another model evaluates the responses generated by the primary system. Developers also use benchmark testing, human review loops, and ranking systems to assess quality.

Evaluation remains one of the least solved yet most critical areas in AI engineering. As companies deploy more AI systems into production, professionals who understand evaluation methodologies will become increasingly valuable.

7. Multimodal AI

AI is no longer limited to text.

Modern AI systems can now process images, videos, documents, screenshots, diagrams, and audio inputs alongside traditional text interactions. This shift toward multimodal AI is reshaping how humans interact with machines.

Today’s models can analyze visual data, extract information from documents, interpret diagrams, and even generate media content. Industries such as healthcare, education, design, and autonomous systems are already benefiting from these capabilities.

Hari noted that while most current commercial applications remain text-focused, multimodal systems represent the next major evolution in AI interfaces. Continue reading on the Packt Medium page.

Data Science & ML Research Roundup

◾How Amazon Finance streamlines regulatory inquiries by using generative AI on AWS: Amazon FinTech teams built a scalable AI-powered regulatory response system using Amazon Bedrock, Claude Sonnet 4.5, OpenSearch, and AWS Lambda. The platform automates document retrieval, multi-turn conversations, and compliance workflows across thousands of regulatory inquiries. With RAG pipelines, vector search, and observability through Langfuse and OpenTelemetry, the system improves accuracy, scalability, security, and response speed for complex compliance operations.

◾Build a Hybrid-Memory Autonomous Agent with Modular Architecture and Tool Dispatch Using OpenAI: This tutorial walks through building a hybrid-memory autonomous AI agent using OpenAI, BM25, vector search, and modular tool orchestration. The architecture combines long-term memory, semantic retrieval, tool calling, and autonomous reasoning into a production-ready framework. With runtime tool swapping, multi-turn memory, and RAG workflows, the system demonstrates how AI agents can reason, remember, and act with scalable autonomy.

◾Automate schema generation for intelligent document processing: AWS has introduced multi-document discovery for the IDP Accelerator, helping teams classify unknown document collections and generate extraction schemas automatically. Using Amazon Bedrock, Cohere Embed v4, k-means clustering, and Strands Agents, the feature groups documents by visual structure, creates JSON schemas, and flags overlaps through quality reports, reducing manual setup for large-scale intelligent document processing.

◾MedAIBase/AntAngelMed: AntAngelMed is a 100B open-source medical LLM built on a highly efficient MoE architecture, activating just 6.1B parameters while matching ~40B dense model performance. Developed by MedAIBase and healthcare partners, it leads benchmarks like HealthBench and MedBench, offering advanced diagnostic reasoning, safety-focused medical intelligence, 128K context support, and high-speed inference exceeding 200 tokens per second.

◾Daybreak | OpenAI for cybersecurity: OpenAI Daybreak is a cybersecurity initiative designed to embed AI-driven defense directly into software development workflows. Combining frontier OpenAI models with Codex Security, the platform helps teams identify vulnerabilities, validate patches, automate remediation, and improve threat modeling at scale. With tiered cyber-focused GPT-5.5 access levels and partnerships across major security firms, Daybreak aims to make software resilient by design through continuous AI-assisted defense.

◾Sparser, Faster, Lighter Transformer Language Models: Sakana AI and NVIDIA introduced TwELL, a sparse data format and custom CUDA kernels that make transformer LLMs faster, lighter, and more energy efficient. By leveraging activation sparsity in feedforward layers, the system delivers over 20% speedups in inference and training while reducing memory usage and power consumption. The research highlights how sparse architectures could become a key path for scaling future LLM performance efficiently.

◾How BASF manages thousands of supply chain decisions with AlphaEvolve: BASF Agricultural Solutions is using Google Cloud’s AlphaEvolve to build a digital twin of its global supply chain, helping planners manage over 5,000 value chains and complex production dependencies. By evolving algorithms from historical operational data, the system improved planning accuracy by over 80%, enabling smarter inventory management, dynamic safety stocks, and network-wide optimization across BASF’s global manufacturing operations.

◾Cloud Storage Rapid turbocharges object storage for AI, analytics: Google Cloud introduced Cloud Storage Rapid, a high-performance object storage family built for AI and analytics workloads. Featuring Rapid Bucket and Rapid Cache, the platform delivers ultra-low latency, multi-terabyte throughput, faster checkpointing, and improved GPU utilization for large-scale AI training and inference. Designed for data-intensive workloads, it helps organizations reduce bottlenecks, lower infrastructure costs, and scale AI systems more efficiently.

◾Building web search-enabled agents with Strands and Exa: AWS and Exa demonstrated how to build web search-enabled AI agents using the Strands Agents SDK and Exa’s AI-native search tools. The integration enables agents to perform semantic web search, retrieve structured page content, and autonomously conduct multi-step research workflows. By combining Exa’s search capabilities with Strands’ model-driven orchestration, developers can build research, fact-checking, and competitive intelligence agents grounded in real-time web data.

See you next time!

Welcome to DataPro #171 — where we unpack the systems, workflows, and AI skills powering production AI today, featuring AI Skills Conf by Community Sprints.

This week, we highlight insights from Richa Awasthi, Vice President at JPMorgan Chase & Co., on how predictive risk analytics is transforming financial stability, helping institutions shift from reactive risk management to real-time, AI-driven decision-making.

Meanwhile, AI is moving far beyond basic prompting, with companies prioritizing skills in agentic workflows, automation, governance, and production deployment. That’s why Packt DataPro is joining the AI Skills Conf on May 14 as a Knowledge Partner, helping bring more practitioner-focused AI learning to our readers alongside the hands-on technical deep dives from the Packt Virtual Conference team.

🚀 Join the Free Virtual AI Skills Conf — May 14

5,000+ professionals | 20+ Speakers from Google DeepMind, AWS, Meta, Spotify, DoorDash, SAP, Scale AI & more

Learn the practical AI skills shaping 2026:

• Agentic workflows and AI-native operations
• AI memory systems and context engineering
• Enterprise AI adoption and governance
• AI ROI and real-world implementation strategies
• Production workflows teams are deploying today

I’ll also be joining the “AI ROI Reality Check” panel alongside leaders from Spotify and SAP.

Register free and secure your spot

Also in this edition:

• Google launches 50+ managed MCP servers for enterprise AI agents
• AWS introduces autonomous frontier agents for DevOps and security operations
• Hapag-Lloyd automates customer intelligence workflows using Amazon Bedrock
• Zyphra releases ZAYA1-8B, a lightweight reasoning MoE rivaling frontier models
• Mend introduces an enterprise AI governance framework for secure AI adoption
• CopilotKit brings persistent memory infrastructure to agentic applications

The next generation of AI work is already taking shape. The advantage now belongs to those learning how these systems actually operate in production.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Reimagining Financial Stability Through Predictive Risk Analytics

By Richa Awasthi, Vice President at JPMorgan Chase & Co.

The real imperative now is not simply to react to risks as they arise, but to anticipate and address them before they can impact the system, leveraging analytical foresight at every step. In this article, I explore how predictive risk analytics can transform not only institutional decision-making, but also the resilience of the broader US financial system.

A turning point is underway in the world of financial stability; institutions are not only shifting their tools but also their mindsets. The introduction and refinement of predictive risk analytics is reshaping practices far beyond mere credit scoring or regulatory compliance. Instead, it is promising a systemic transformation; one that’s anchored in real-time data, advanced machine learning, and an unrelenting drive to foresee threats before they trigger a crisis.

‘By integrating predictive analytics into every layer of financial oversight, these innovations are reimagining how systemic risk is managed in the US, ensuring a more resilient and inclusive economy.’

The Evolving Landscape of Financial Stability

Over the past decade, US financial institutions have navigated significant economic turbulence, including volatile interest rates, unexpected inflationary pressures, and the emergence of new asset classes — factors that have exposed the limitations of traditional frameworks for managing systemic risk.

According to the Federal Reserve and the United States Economic Forecast by Deloitte, the average CPI growth in 2025 is projected at 2.9%, moderating in 2026 before falling closer to 2.3% by the end of the decade. Meanwhile, the baseline forecast for the 10-year Treasury yield suggests sustained pressure, remaining above 4.1% through 2030, even as short-term rates fluctuate in response to shifting monetary policies.

This emphasises, ‘Strengthening US financial stability requires embedding predictive insights across all layers of risk oversight, from credit allocation to liquidity management, so that institutions can anticipate emerging threats, respond proactively, and maintain systemic resilience.’

The Rise of Predictive Analytics

The predictive analytics market is currently experiencing explosive growth. A 2024 report by MarketsandMarkets projects the global market for predictive analytics will climb from $10.5 billion in 2023 to $14.5 billion by 2024, registering an impressive 13.5% CAGR. Fueled by AI and machine learning, these systems now comb through vast, multidimensional datasets, ranging from transaction histories to alternative data sources, identifying patterns that are invisible to conventional models.

Risk management is the primary beneficiary. Financial institutions are leveraging predictive models that assess real-time exposures, estimate macroeconomic sensitivities, and recalibrate risk profiles in response to destabilising events. According to PwC, by 2030, 95% of financial models will incorporate Environmental, Social, and Governance (ESG) factors, reflecting a broader commitment to holistic oversight.

‘Preventing banking crises is no longer a matter of luck or last-minute intervention. Predictive analytics offers us a toolkit for ongoing surveillance, early warning, and targeted mitigation, fundamentally rewiring the calculus of risk.’

Transforming Small Business Access to Credit

While the macroeconomic benefits of predictive risk analytics are clear, its implications for small business lending are particularly consequential for US economic growth and resilience.

Traditionally, access to credit relied on rigid, historical measures, often excluding startups, minority entrepreneurs, and those with unconventional financial profiles. Predictive analytics, by contrast, enables lenders to assess future potential, not merely past performance, allowing capital to flow where it can generate sustainable economic impact.

Recent analyses indicate that global financial institutions leveraging predictive models have reduced business loan default rates by up to 30%, while approval rates for previously underserved applicants increased by over 25%.

Predictive Models as Crisis Prevention Tools

Financial crises have frequently exposed the limitations of traditional risk management. The 2008 banking crisis, for instance, exposed the fragility of models overly reliant on lagging indicators and static assumptions.

Contemporary predictive models utilise statistical, machine learning, and data mining techniques to identify risks well in advance of their materialisation. Early interventions allow for targeted responses rather than blanket interventions.

‘Banking crises don’t happen overnight; they simmer. What predictive analytics allows us to do is notice when the system is beginning to heat up and take action while there’s still time to avert damage.’

A Note of Caution: The Human Factor

While predictive analytics can help illuminate systemic weak spots, there remains a real danger in over-reliance on algorithmic models. Human judgment, ethical reasoning, and holistic context are still critical.

‘Predictive analytics must complement, not supplant, prudent oversight and ethical grounding.’

Technology, Policy, and the Path Ahead

Policy initiatives across the US are increasingly focused on harnessing such tools for broader stability. The Federal Reserve plans to continue integrating advanced analytics into its supervisory frameworks.

Reflecting Forward: A System Reimagined

As the decade unfolds, the convergence of predictive analytics, policy innovation, and professional expertise points toward a financial system both more resilient and more inclusive.

‘Predictive analytics is a critical tool, yet the real transformation occurs when we rebuild processes and organisational culture around its insights.’

Data Science & ML Research Roundup

◾ Google-managed MCP servers are available for everyone: Google Cloud announced 50+ managed MCP servers for AI agents, enabling secure, enterprise-grade connectivity across Google Cloud services. These servers support interoperability with tools like ChatGPT, Claude, LangChain, and Gemini CLI while offering centralized discovery, governance, observability, and security. Use cases span infrastructure automation, analytics, developer assistance, productivity workflows, and customer experiences, helping agents move from prototypes to production-ready autonomous systems.

◾ AWS launches frontier agents for security testing and cloud operations: AWS announced two new frontier AI agents: AWS Security Agent and AWS DevOps Agent, designed for autonomous security testing and cloud operations. AWS Security Agent reduces penetration testing from weeks to hours by autonomously identifying and validating vulnerabilities. AWS DevOps Agent accelerates incident resolution across multicloud environments with up to 75% lower MTTR. Together, they represent AI systems that independently manage complex, persistent enterprise workflows.

◾ How Hapag-Lloyd uses Amazon Bedrock to transform customer feedback into actionable insights: Hapag-Lloyd built an AI-powered customer feedback analysis platform using Amazon Bedrock, OpenSearch, LangChain, and LangGraph to automate sentiment analysis, insight generation, and reporting. The solution replaces manual feedback reviews with scalable AI workflows, enabling faster product decisions and real-time insights. Features include AI chatbots, biweekly reports, semantic search, and Bedrock Guardrails for responsible AI, helping teams focus more on innovation and customer experience.

◾ Zyphra Releases ZAYA1-8B: A Reasoning MoE Trained on AMD Hardware That Punches Far Above Its Weight Class: Zyphra released ZAYA1-8B, an Apache 2.0 open-weight Mixture-of-Experts model with 760M active and 8.4B total parameters, trained entirely on AMD MI300 hardware. Despite its small active footprint, it rivals frontier reasoning models in math and coding benchmarks using innovations like MoE++, 8× KV-cache compression, and Markovian RSA test-time compute, enabling efficient, high-performance reasoning with lower inference costs and latency.

◾ Mend Releases AI Security Governance Framework: Covering Asset Inventory, Risk Tiering, AI Supply Chain Security, and Maturity Model: Mend released an AI Security Governance Framework to help organizations manage rapidly growing AI adoption through structured governance, risk tiering, and supply chain security. The framework covers AI asset inventory, least-privilege access, AI-BOMs, monitoring for AI-specific threats, and maturity modeling aligned with NIST, OWASP, ISO, and EU AI Act standards. It provides practical guidance for scaling AI securely without slowing engineering velocity.

◾ CopilotKit Introduces Enterprise Intelligence Platform That Gives Agentic Applications Persistent Memory Across Sessions and Devices: CopilotKit introduced its Enterprise Intelligence Platform, adding persistent memory infrastructure for agentic applications across sessions and devices. Built on top of the CopilotKit stack, the platform enables durable “Threads” that preserve UI state, workflows, voice, files, and multimodal interactions. It supports framework-agnostic agents, enterprise security features, and resumable long-running workflows, helping teams move agentic applications from stateless demos to production-ready systems.

See you next time!

Hello there!

Welcome to DataPro #170

This week, we go deeper into AI systems, not just how they perform, but how they behave, and why. We’re featuring an editorial from our Content Engineer, Gowri Rekha, based on A Practical Guide to Reinforcement Learning from Human Feedback. Her core idea: alignment isn’t added later, it defines the system. AI systems don’t fail because they misunderstand intent. They fail because they optimize exactly what they’re given. The real question is: what should a system optimize for?

You’ll see the same challenge in forecasting, where model choice and evaluation shape outcomes. If you work with time series, join our live workshop on May 2 with Jeffrey Tackes and Manu Joseph to build and benchmark a full forecasting pipeline. Lock in your seat with 35% off.

Looking ahead,

On May 14, the AI Skills Conf brings together 30+ speakers from Meta, Google, AWS, Scale AI, Bolt, and more to focus on real-world AI use cases, adoption decisions, and what the AI stack actually looks like in practice.

Expect sessions on:
– The AI skills professionals need heading into 2026
– How companies are actually deciding which tools to adopt
– What a real-world AI stack looks like for teams and founders

I’ll be joining a panel on “AI ROI reality check”, unpacking where AI is delivering measurable value and where it still falls short. It’s free to attend, register now to secure your spot.

Register for free

This is part of our effort as a knowledge partner with global AI communities to bring more practitioner-focused learning to our readers, complementing the in-depth, hands-on technical sessions from our Packt Virtual Conference team.

Alongside this, our research roundup brings together some of the most important movements shaping modern AI systems right now.

This week’s highlights:

• GitNexus reframes coding agents by giving them structural awareness of codebases

• Vision Banana blurs the line between generation and perception in computer vision

• Amazon Quick and Visier show how agentic systems are entering enterprise decision workflows

• Gemini Agent Platform signals the shift toward governed, production-grade agent ecosystems

• SageMaker automates one of the hardest steps in GenAI: deployment optimization

• Sapiens2 pushes human-centric vision to new levels of detail and realism

• Grok Voice moves voice AI closer to real-time, interruptible conversations

• DeepSeek-V4 expands what open models can do with long context and efficiency

• OpenAI’s Privacy Filter brings practical, scalable PII detection into pipelines

• Qwen3.6 focuses on making coding agents actually usable in real-world workflows

If there’s a common thread, it’s this: AI is shifting from isolated models to systems that operate in context. In that shift, what we optimize for, how we connect data, and how we act on it are becoming just as important as the models themselves.

Let’s get into it.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

AI Alignment Isn’t a Fix—It’s the Whole System

Why RLHF and modern alignment methods are really about defining what AI systems optimize for?

By Gowri Rekha, Content Engineer, Data Science.

There is a prevalent assumption in how we talk about artificial intelligence: that systems are built first and aligned later. In this view, alignment behaves like a finishing layer, something applied once the real engineering work is done. It is an appealing idea because it suggests that intelligence and intention can be handled separately.

The truth though is that they cannot be handled separately.

AI does not understand in the way people assume. They do not reason about intent or meaning in a human sense. Instead, they learn patterns from data and feedback, and optimize for the signals they are given. They don’t always get it right, of course, but they reliably move in the direction they’re rewarded. That is both their strength and their risk. When a system is designed to optimise, it will do so. Even if the objective itself is poorly defined.

This is why alignment is not a correction. The real issue is not that AI systems fail to do what they are told. It is that they succeed a little too well. Behaviour is shaped by what systems are trained to optimize. This is true across AI systems, but it is easiest to see in reinforcement learning, where what gets rewarded becomes more likely. The challenge is defining what “desirable” actually means.

Human goals, just like human desires, are rarely precise. They are shaped by context, judgment and sometimes contradiction. We rely on intuition and shared understanding. These do not translate easily into metrics and reward functions. When they are reduced into simplified metrics, systems begin to exploit the gaps. What emerges is not failure, but a kind of distorted success. The system achieves the objective as defined, while missing its intent. This is often called reward hacking, though there is nothing accidental about it. It is the natural outcome of a system doing exactly what it has been asked to do.

This is where the idea of alignment as an afterthought breaks down. Alignment cannot be added later because it sits at the point where success is defined. It exists before the model, before the training loop, before the data pipeline. Once a system is trained to optimize for signals derived from data, human feedback, or reward models, those signals shape the behavior that emerges. Trying to fix this later is like changing performance metrics in a company that has already optimised around the wrong ones. The issue, you see, is not reporting. It is the design itself.

One practical response has been to involve humans directly in the learning process. Instead of trying to define “good” in advance, systems are guided by human preferences. People compare outputs, make judgments and provide corrections. Reinforcement learning from human feedback, or RLHF, formalises this approach by turning subjective evaluation into a training signal.

It works not because it is perfect, but because it accepts a simple reality. Alignment is not only a technical problem. It is a human one.

Introducing humans, however, brings its own complications. Feedback is inconsistent. It varies with context, bias and fatigue. In effect, it trades one set of challenges for another . This has led to newer methods that try to reduce reliance on direct human input and improve scalability while retaining its intent. Techniques such as Reinforcement Learning from AI Feedback and Constitutional AI are part of this shift. Contrastingly, DPO uses human data but simplifies the process by removing the need to have a reward model.

As the author, Sandip Kulkarni, of A Practical Guide to Reinforcement Learning from Human Feedback explains: “These methods reduce cost and complexity. RLAIF lowers human effort, DPO removes reward modeling, and Constitutional AI adds structured principles. They are useful in specific scenarios, but RLHF still performs better in some cases, especially complex reasoning or high-stakes domains. None fully replace RLHF yet, but are important for their strengths.”

What is emerging is not a replacement for RLHF, but a broader toolkit. Different methods suit different constraints, and alignment is becoming less about a single technique and more about choosing the right approach for the context.

Understanding these ideas is one challenge. Building intuition around them is another. Much of the difficulty comes from the way alignment is presented, often as something abstract or detached from core machine learning concepts. Sandip takes a different approach: “The book starts by introducing reinforcement learning concepts and relating RLHF and transfer learning to core algorithms like Q-learning. It demonstrates how human feedback can be integrated in simplified RL environments, such as gridworld, a classical navigation problem that reduces cognitive load and makes the underlying mechanics easier to understand. This helps readers see RLHF as part of a broader AI system rather than just a recipe or isolated algorithm.”

The progression from simple environments to large language models reflects a more grounded way of thinking about alignment. It is not an isolated technique applied at scale, but an extension of fundamental reinforcement learning principles. By building this foundational base, readers can see how advanced techniques like Policy Gradients, GAE, and PPO are not just isolated recipes, but a natural scaling of these fundamental principles to the world of Large Language Models.

The motivation behind Sandip Kulkarni’s work also points to a larger gap in the field: “RL and LLMs are tools and methodologies that enable today’s AI capabilities. While the engineering processes for these are relatively well defined, and continue to show improvements, the AI alignment problem has been a critical frontier problem at the intersection of the Human-AI interface. I didn’t find literature bridging RL, LLMs, and AI alignment in the way I was looking for, so I decided to do it myself.”

This gap between capability and control is where much of the current progress in AI is concentrated. The conversation is shifting from how to build more capable systems to how to ensure those systems behave in ways that are consistent with human intent.

The implication is straightforward. If alignment is treated as secondary, improvements in capability will amplify existing problems. If it is treated as foundational, capability and behaviour evolve together.

In that sense, alignment is not a feature of AI systems. It is the system. It determines not only what they can do, but how they will behave when given the opportunity.

For those working in this space, the challenge is not just to understand alignment, but to apply it. A Practical Guide to Reinforcement Learning from Human Feedback is designed to bridge that gap, connecting core reinforcement learning ideas with practical alignment techniques across different scales.

Read more here in the book, A Practical Guide to Reinforcement Learning from Human Feedback.

We’re hosting a free live session on DeerFlow, where core contributors will walk through the open-source SuperAgent framework for building autonomous, long-running AI workflows—join us on May 6.

Data Science & ML Research Roundup

◾Meet GitNexus: An Open-Source MCP-Native Knowledge Graph Engine That Gives Claude Code and Cursor Full Codebase Structural Awareness GitNexus tackles a core flaw in AI coding agents: lack of codebase awareness. It builds a full knowledge graph of repositories, mapping dependencies, execution flows, and structure, then serves it via MCP to agents. This enables accurate impact analysis, safer edits, and architectural reasoning, turning AI from guesswork into informed decision-making, even for smaller models.

◾Google DeepMind Introduces Vision Banana: An Instruction-Tuned Image Generator That Beats SAM 3 on Segmentation and Depth Anything V3 on Metric Depth Estimation. Google’s Vision Banana challenges a core assumption in computer vision: generative models can also understand images. Built by lightly tuning an image generator, it matches or beats specialist models across segmentation, depth, and normals, without task-specific architectures. By framing perception as image generation, it turns one model into a generalist vision system, signaling a shift toward unified, foundation-level visual intelligence.

◾Building Workforce AI Agents with Visier and Amazon Quick: Amazon Quick and Visier combine to solve a core enterprise gap: fragmented decision-making. By connecting workforce intelligence with organizational context through MCP, they create a unified agentic workspace where users can query, analyze, and act in one place. The result is faster insights, automated workflows, and context-rich decisions grounded in both live data and internal policies.

◾Amazon Quick for marketing: From scattered data to strategic action. Amazon Quick tackles a core marketing bottleneck: disconnected tools and scattered insights. By unifying data across systems into a personal knowledge graph, it delivers real-time campaign intelligence, automated research, and scalable content creation. The result is faster execution, automated workflows, and a shift from manual reporting to strategy, where insights are instantly actionable.

◾Introducing Gemini Enterprise Agent Platform: Google’s Gemini Enterprise Agent Platform reframes how enterprises build and manage AI agents. It brings model access, development tooling, orchestration, and governance into one system. It builds on Vertex AI but targets a more complex, multi-agent world. Teams can design, deploy, and coordinate agents across systems. Built-in memory, security, and observability ensure agents operate with context and control at scale.

◾Amazon SageMaker AI now supports optimized generative AI inference recommendations: Amazon SageMaker AI rethinks one of the most painful parts of GenAI: getting models into production. Instead of trial-and-error benchmarking, it automates the entire process, from narrowing infrastructure choices to applying the right optimizations and validating them on real GPUs. What used to take weeks of tuning and guesswork becomes a guided, metrics-driven path to production readiness.

◾Meta AI Releases Sapiens2: A High-Resolution Human-Centric Vision Model for Pose, Segmentation, Normals, Pointmap, and Albedo. Meta’s Sapiens2 goes after a long-standing gap in vision models: truly understanding humans. It blends reconstruction and contrastive learning to capture both texture and meaning, trained on a massive, carefully curated human dataset. Designed for high-resolution reasoning, it advances how models interpret pose, body structure, and surface details across real-world conditions.

◾xAI Launches grok-voice-think-fast-1.0: Topping τ-voice Bench at 67.3%, Outperforming Gemini, GPT Realtime, and More. xAI’s grok-voice-think-fast-1.0 pushes voice AI closer to real conversations. It handles full-duplex interaction, interruptions, and noisy inputs while reasoning in real time without latency tradeoffs. Built for complex workflows, it captures structured data accurately and operates across languages, with strong benchmark performance and proven scale in live customer support and sales environments.

◾deepseek-ai/DeepSeek-V4-Pro · DeepSeek-V4 pushes open-source LLMs into long-context and high-efficiency territory. Built as a Mixture-of-Experts system with up to 1M token context, it combines hybrid attention and optimized training to reduce compute while scaling capability. With flexible reasoning modes and

strong coding, reasoning, and agent benchmarks, it closes the gap with leading closed models.

◾openai/privacy-filter · OpenAI’s Privacy Filter focuses on a practical gap in AI pipelines: reliable PII detection at scale. It uses a bidirectional token-classification approach to identify and mask sensitive data in a single pass. Lightweight, fast, and deployable on-prem, it’s designed for high-throughput workflows where teams need controllable, context-aware privacy filtering.

◾unsloth/Qwen3.6-27B-GGUF · Qwen3.6 focuses on making coding LLMs more usable in real workflows. It improves agentic coding, repository-level reasoning, and tool calling, while preserving context across interactions. Designed with developer feedback, it supports long contexts and integrates easily with modern inference stacks, aiming to deliver a more stable and practical coding assistant experience.

See you next time!

Thesys Agent Builder lets you build an AI Agent that turns your data into stunning, interactive UI in real-time. No more walls of text, just visual, actionable insights right inside the conversation.

Connect databases like Snowflake, Databricks and more. Describe what the agent does and deploy. Your agent is live within minutes! No SQL, no code, no new infra needed. Now your agent handles natural language questions and responds with interactive charts, tables, and dashboards instead of text.

Try it for free

Hello there!

Welcome to DataPro 169.

This week’s issue sits at the intersection of performance, control, and scale across the modern data and AI stack. As systems grow more autonomous and multimodal, the underlying question remains the same: how do we design systems that are both efficient and reliable in production?

We start with an expert-led deep dive from Packt author Jun Shan on PostgreSQL indexing, a foundational capability that determines whether your data systems scale gracefully or collapse under real-world workloads. It’s a reminder that even as AI evolves rapidly, core data infrastructure decisions still shape performance outcomes.

Across the rest of the issue, that same theme shows up in different forms: models becoming more customizable, agents becoming more autonomous, infrastructure becoming more scalable, and governance becoming more critical. From fine-tuned LLMs and self-optimizing agents to multimodal generation and secure AI deployment, the stack is evolving toward systems that can both act independently and be controlled precisely.

Here’s what else is shaping the data and AI landscape this week:

• Fine-tuning Nova models on Amazon Bedrock is making domain-specific AI faster, cheaper, and more accurate

• Google’s Lyria 3 models bring controllable, studio-quality music generation into applications

• AutoAgent shows how AI can now optimize its own agent workflows without human tuning

• Project Glasswing signals a shift toward AI-powered cybersecurity defense at scale

• Gemma 4 expands open multimodal models with strong reasoning and on-device capabilities

• OSGym makes large-scale computer-use agent training dramatically more affordable

• Human-in-the-loop patterns are becoming essential for safe AI deployment in healthcare

Let’s dive in.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Design production-ready Power BI and Fabric analytics systems that scale across teams and complexity, with governance built in from day one. Get a complete architecture playbook, capacity planning tools, semantic modeling framework, real-world case studies, and a free bestselling Power BI eBook.

Save 35% with code ANALYTICS35 🎟

PostgreSQL Indexing

By Jun Shan

Introduction

Indexes are one of the most important performance tools in the PostgreSQL databases. Without indexes, even a well-designed schema can become slow as data grows. Many systems work perfectly during development, when tables contain thousands of rows. But in production, tables often grow to millions or hundreds of millions of rows. At that scale, queries that once took milliseconds can suddenly take minutes or even hours.

Consider a customer support system where agents need to look up orders by customer ID. If the orders table contains 50 million records and there is no index, PostgreSQL must scan every row to find matching orders. This is called a sequential scan. Even on fast hardware, scanning millions of rows repeatedly can overwhelm the system and slow down the entire application.

Proper indexing is often the difference between a system that scales and one that fails under real-world load. The author has worked with production database systems where indexing reduced query time from several minutes to milliseconds. In this blog, we will go through some introductory understanding of PostgreSQL indexes.

The Core Concept

An index is a supporting data structure that helps PostgreSQL find rows faster. It works much like an index in a book. Instead of reading every page to find a topic, you use the index to jump directly to the right location. Indexes matter because they dramatically reduce the amount of work required to answer queries. Without an index, PostgreSQL must perform a sequential scan. This means reading every row in the table and checking whether it matches the query condition. With an index, PostgreSQL can locate the relevant rows quickly, often by examining only a small portion of the data.

The most common type of index in PostgreSQL is the B-tree index. It organizes values in sorted order, which allows fast lookup, filtering, and sorting. When you create an index on a column, PostgreSQL builds a structure that allows efficient searching based on that column’s values.

Indexes are powerful, but they must be used strategically. Many practitioners misunderstand how indexes work. One common mistake is assuming that indexes automatically improve every query. In reality, indexes help only when the query uses the indexed column in filtering, joining, or sorting. Another mistake is creating too many indexes. Each index consumes storage and slows down insert, update, and delete operations because PostgreSQL must maintain the index structure.

Subscribe now

Practical Example

Consider an e-commerce system with a table that stores customer orders:

orders

------

order_id

customer_id

order_date

status

total_amount

A common query might look like this:

SELECT *

FROM orders

WHERE customer_id = 1001;

If the table contains millions of rows and there is no index, PostgreSQL must scan every row to find orders for customer 1001. This can take a long time.

To fix this, you can create a B-tree index:

CREATE INDEX idx_orders_customer_id

ON orders(customer_id);

This creates a sorted structure based on customer_id. Now, when PostgreSQL runs the query, it can use the index to locate matching rows quickly instead of scanning the entire table.

The improvement can be dramatic. A query that previously took several seconds might now complete in a few milliseconds.

Indexes also help with range queries. For example:

SELECT *

FROM orders

WHERE customer_id IN (1001, 1003, 1007)

AND order_date = ‘2026-03-01’;

If you create a composite index:

CREATE INDEX idx_orders_customer_date

ON orders(customer_id, order_date DESC);

PostgreSQL can query with both filters efficiently using the same index. This is especially valuable in applications where users frequently view recent orders, transactions, or events.

Common Pitfalls

While indexes are powerful, improper use can reduce performance instead of improving it.

One common mistake is over-indexing. Each index must be updated whenever data changes. If a table has too many indexes, insert and update operations become slower. This is because PostgreSQL must update every related index in addition to the table itself.

Another mistake is indexing columns with low selectivity. Selectivity refers to how unique the values are. For example, indexing a column with only two values, such as status (“active” or “inactive”), may not help much. PostgreSQL may still choose a sequential scan inside each status’ storage because the index does not narrow down the results significantly.

Composite indexes introduce another challenge: column order matters. For example, an index on (customer_id, order_date) works well for queries filtering by customer_id and order_date, or by customer_id alone. But it may not help queries filtering only by order_date. Understanding query patterns is critical when designing composite indexes.

Many experienced users also forget to analyze query execution plans. PostgreSQL provides the EXPLAIN and EXPLAIN ANALYZE commands, which show whether an index is being used. Simply creating an index does not guarantee PostgreSQL will use it.

Another advanced consideration is index maintenance. PostgreSQL uses a process called VACUUM to clean up dead rows. Without proper maintenance, indexes can become bloated and inefficient. Regular VACUUM and ANALYZE operations help keep indexes effective.

Some of the most effective optimization strategies include:

• Indexing columns used in WHERE, JOIN, and ORDER BY clauses

• Using composite indexes for frequent multi-column queries

• Avoiding unnecessary indexes

• Monitoring query performance regularly

• Reviewing execution plans to confirm index usage

Good indexing is not about creating many indexes. It is about creating the right indexes.

Subscribe now

Conclusion

Indexes are essential for making PostgreSQL databases fast and scalable. They allow PostgreSQL to locate data efficiently without scanning entire tables. This becomes increasingly important as databases grow in size and complexity.

The key insight is that indexes are not automatic performance boosters. They must be carefully designed based on real query patterns. The most effective indexes target frequently used filters, joins, and sorting operations. At the same time, excessive or poorly designed indexes can slow down write operations and waste resources.

The broader impact of proper indexing is significant. It improves application responsiveness, reduces server load, and enables systems to scale to millions or billions of records. In many real-world systems, indexing is the single most important factor in database performance.

If you work with PostgreSQL, reviewing your indexes is one of the highest-value improvements you can make. Use tools like EXPLAIN ANALYZE, observe real query behavior, and design indexes intentionally. Even small changes can produce dramatic performance gains.

For further information, especially the behind scene definitions of how different types of indexes are created and utilized, you can refer to Chapter 10, Performant SQL in my book, SQL for Data Analytics, 4th Edition, where we provide comprehensive coverage to PostgreSQL indexes. Your feedback on this topic is greatly appreciated.

Solve real-world data problems by combining SQL with GenAI in a hands-on session built for practical workflows and faster insights. Walk away with reusable frameworks, prompt templates, real datasets, debugging checklists, and a free bestselling SQL eBook.

Save 35% with code GENAI35 🎟

Data Science & ML Research Roundup

◾ Customize Amazon Nova models with Amazon Bedrock fine-tuning. To customize AI models effectively, Amazon Bedrock enables fine-tuning of Amazon Nova models by embedding domain-specific knowledge directly into the model, improving accuracy, reducing latency, and lowering costs; this article explains how Bedrock supports customization through supervised and reinforcement fine-tuning and distillation, compares these methods with prompt-based approaches, and demonstrates their impact using an intent classification example that significantly boosts performance.

◾ Lyria 3 and Lyria 3 Pro on Vertex AI: To generate high-quality music for applications, Google’s Lyria 3 models on Vertex AI enable audio creation from text and images with vocals, lyrics, and structured compositions; this is about new generative music models that offer flexible track lengths, creative control, and commercial safety, helping businesses integrate studio-quality, predictable audio production into apps, workflows, and content pipelines.

◾ Meet ‘AutoAgent’: The Open-Source Library That Lets an AI Engineer and Optimize Its Own Agent Harness Overnight. AutoAgent is an open-source library that enables a meta-agent to autonomously optimize its own agent harness without human intervention; this is about replacing manual prompt-tuning loops with AI-driven iteration on prompts, tools, and orchestration, achieving top benchmark results and shifting the engineer’s role from hands-on optimization to setting high-level goals.

◾Project Glasswing: Securing critical software for the AI era. Project Glasswing is a collaborative initiative bringing together major tech and security organizations to use advanced AI models like Anthropic’s Claude Mythos Preview for cybersecurity defense; this is about addressing the growing risk of AI-driven vulnerability discovery by deploying frontier models to proactively identify and secure critical software, while scaling industry-wide efforts to stay ahead of emerging threats.

◾ google/gemma-4-31B-it: Gemma 4 is Google DeepMind’s latest family of open multimodal models designed for efficient, high-performance AI across devices; this is about scalable models that support text, images, audio, and video with strong reasoning, coding, and agent capabilities, offering flexible architectures and deployment options from on-device environments to high-performance systems.

◾ Meet OSGym: A New OS Infrastructure Framework That Manages 1,000+ Replicas at $0.23/Day for Computer Use Agent Research. OSGym is an open-source infrastructure framework designed to make training computer-use AI agents scalable and affordable; this is about solving the high cost and complexity of running thousands of OS environments by optimizing orchestration, storage, and fault recovery, enabling large-scale agent training at dramatically lower costs for research and development.

◾ Human-in-the-loop constructs for agentic workflows in healthcare and life sciences. Human-in-the-loop (HITL) constructs are essential for safely deploying AI agents in healthcare and life sciences; this is about integrating human oversight into agent workflows using AWS tools to ensure regulatory compliance, patient safety, and auditability, while maintaining automation efficiency through patterns like approval hooks, tool-level controls, asynchronous workflows, and real-time elicitation.

See you next time!

Welcome to DataPro 168, and this week we’re looking at a shift that feels bigger than any single model launch or product update: the stack is maturing, fast.

AI is no longer just about smarter models. It is about resilient infrastructure, better deployment economics, multimodal interfaces, self-improving agents, and systems that can actually survive production reality.

That’s why this week’s deep dive article feels especially relevant. In this expert-led piece, Stéphane Eyskens, Senior Cloud & Cloud Native Architect (Cloud Native and iPaaS MVP), cuts through the abstraction to get to the operational reality: performance means very little if your architecture cannot hold up when a region fails.

In Data Resilience in Azure for OLTP Data Stores, he breaks down how Azure SQL, Managed Instance, Cosmos DB, DocumentDB, and Storage behave under real failure scenarios, mapping the actual trade-offs behind RTO, RPO, failover design, and cross-region networking.

In a week full of noise about what AI can do, this is the kind of grounded analysis that reminds us what production systems still have to guarantee.

Because in the end, the future does not belong to the most impressive demo. It belongs to the systems that can reason, recover, scale, and stay available when it matters.

🔍 This week’s highlights

• Bedrock RFT goes practical
  Feedback-driven fine-tuning with smaller datasets and OpenAI-style APIs.

• Vertex AI resilience patterns
  How to handle 429s with retries, routing, and smarter traffic design.

• Gemini Live Agent Challenge
  Build multimodal agents, win up to $80K.

• SageMaker GPU reservations
  Predictable inference capacity with training plans.

• Hyperagents
  Self-improving AI that rewrites its own learning logic.

• OpenSpace
  Skill-based agents that cut tokens and improve over time.

• New open models (Nemotron + Qwen)
  Stronger reasoning, agentic behavior, and more control.

Struggling to move from ML experiments to production systems?

This week, join live PyTorch workshop with Ashish Ranjan Jha (Mar 29, 7–10 PM GMT+5 | 9:30 AM–12:30 PM EDT) and build the engineering patterns teams use to ship real ML systems. Save your seat now.

Save Your Seat

Let’s dive in and unpack what it really takes to build systems that hold up under pressure.

See you inside.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Data resilience in Azure for OLTP data stores | by Stephane Eyskens

Azure offers many ways to replicate data and implement multi-region architectures. The purpose of this blob post is to illustrate how to achieve the best possible Recovery Time Objective (RTO) and Recovery Point Objective(RPO) can be achieved using the most commonly used Azure services. I’m leaving out of scope backup/restore and application-level techniques that help build an exhaustive resilience posture.

For each service, two situations are considered:

• Both regions are fully up and running

• The primary region is completely unavailable

The scenarios also assume isolation from the Internet, as this is the most representative setup for enterprise-grade architectures.

If you test these patterns in your own environment, the primary region will likely remain available, which may give the impression that the system behaves differently but make no mistake, this behaves exactly as described below. Let’s start with Azure SQL, one of the main OLTP service in Azure.

Azure SQL

In the diagram, I illustrate how geo-replication and failover groups compare. My preference is for failover groups, as they handle DNS more elegantly through static listener endpoints. They also allow you to fail over one or multiple databases together, which is not the case with standard geo-replication. With Azure SQL Database, the primary and secondary servers automatically discover each other, simplifying the configuration.

Active Geo-Replication

Azure SQL — Data Replication through geo-replication

Failover Groups

Azure SQL — Data Replication using failover groups

Azure SQL Managed Instance

Unlike Azure SQL Database, Managed Instances only support failover groups for geo-replication and rely on the customer’s network infrastructure to carry the replication traffic. Therefore, it is essential to ensure that the primary and secondary instances can reach each other through the underlying network and DNS architecture.

Because this connectivity depends on the customer’s network design, multiple implementation patterns are possible, which all have a positive or negative impact on the RPO.

Replication through the hubs

Managed Instance — Data replication through hubs

In this setup, each managed instance spoke is peered with its respective regional hub, and the regional hubs are also peered with each other. In some organizations, dedicated integration hubs may fulfill this role instead of relying on a single hub for all connectivity.

Routing must be configured so that the data replication path follows: primary → primary regional hub → secondary regional hub → secondary. Firewall rules must be defined accordingly to allow this traffic, and in both directions since the secondary also initiates calls to the primary.

While fully aligned with the hub-and-spoke model, a major drawback of this design is that the replication traffic traverses two firewalls. For highly active databases, this may introduce additional load on the firewalls. If the firewalls do not scale adequately, this could ultimately impact the RPO.

Replication through spokes peering

Managed Instance — Replication through spoke peering

This setup is the simplest, as both Managed Instance spokes are directly peered with each other. No additional configuration is required since the primary and secondary instances can communicate through the peering.

However, the main drawback is that this design does not align with the typical hub-and-spoke topology, where peerings are generally allowed between spokes and hubs only, and not directly between spokes.

Replication through a hybrid approach

Managed Instance — Hybrid Approach

In this setup, both Managed Instance spokes are peered with their respective regional hub and also directly peered with each other. This enables the use of the direct peering for data replication, while all other traffic continues to be routed through the hubs.

The advantage is that replication traffic no longer traverses firewalls, reducing latency and avoiding additional load on them, while firewalls still enforce policies for other traffic flows.

However, this approach again deviates from the traditional hub-and-spoke model, since the two spokes are directly peered. Moreover, because the virtual networks are peered, each VNet can see the entire address space of the other. As a result, a routing misconfiguration on non–Managed Instance subnets could inadvertently allow other traffic to bypass the firewalls.

Replication through subnet-level peering

Managed Instance — Replication through subnet-level peering

An alternative to the previous setup is to use subnet-level peering rather than peering the entire virtual networks. In this model, only the Managed Instance subnets are peered, allowing the primary and secondary instances to communicate directly, while all other subnets in the two virtual networks can only see the address range of their respective regional hub.

The advantage of this approach is that firewalls cannot be bypassed for any traffic unrelated to the Managed Instances, while replication traffic follows the most direct network path with no additional friction.

However, this design still deviates from the traditional hub-and-spoke topology, since the spokes are partially peered with each other, even though the peering is limited to specific subnets.

Replication through virtual hubs (VWAN)

In Azure Virtual WAN, virtual hubs have a default router that let’s all spokes talk to each other automatically. Hub to hub communication is also possible providing inter-hub connectivity is configured. In this case, there is no more firewall on the data replication path.

Cosmos DB single vs multi-region writes

Cosmos DB is well suited for multi-region architectures, as it supports multi-region writes. In addition, the SDKs automatically route requests to available regions, allowing applications to continue operating without any changes on the development side.

In this diagrams, I illustrate both single-write with read-only replicas and multi-region write configurations.

Cosmos single-region write

Cosmos DB — Single write region

Cosmos multi-region write

Cosmos DB — multi-region write

DocumentDB single vs multiple DNS zones

DocumentDB is interesting in terms of DNS behavior since it makes use of SRV records to discover the actual nodes that identify as primary or secondary. Therefore, working with a single or multiple DNS zones makes a big difference.

DocumentDB single DNS zone

DocumentDB — Single DNS Zone

Continue reading this full article on our Packt Medium handle.

Is your Power BI architecture ready to scale?
Design Fabric and Power BI systems that stay fast, governed, and production-ready.

Receive a full architecture playbook, capacity planning worksheet, semantic modeling framework, governance checklist, plus a free Learn Power BI eBook, templates, case studies, and certification. Lock in your seat

Data Science & ML Research Roundup

◾Reinforcement fine-tuning on Amazon Bedrock with OpenAI-Compatible APIs: a technical walkthrough: Reinforcement fine-tuning (RFT) on Amazon Bedrock simplifies LLM customization by learning from feedback, not large datasets, using automated workflows and OpenAI-compatible APIs. This post walks through the full setup, from authentication to Lambda-based reward functions and training jobs, demonstrating how to fine-tune models like GPT-OSS-20B on GSM8K, while explaining RFT’s core concepts, components, and advantages.

◾ HauhauCS/Qwen3.5-35B-A3B-Uncensored-HauhauCS-Aggressive: Qwen3.5-35B-A3B Uncensored (Aggressive) delivers a fully unlocked LLM with zero refusals, preserving original capabilities while enabling unrestricted outputs across text, vision, and reasoning tasks. This post outlines its architecture, MoE design, multimodal support, quantized deployment options, recommended inference settings, and how to run it locally via GGUF-compatible tools like llama.cpp and LM Studio.

◾Meta AI’s New Hyperagents Don’t Just Solve Tasks—They Rewrite the Rules of How They Learn: Hyperagents unlock recursive self-improvement by making both task-solving and self-modification fully editable within a single program, enabling transferable, domain-agnostic learning beyond fixed meta-systems. This post explores how DGM-H overcomes infinite regress, introduces metacognitive self-modification, and demonstrates strong gains across robotics, paper review, and math—while revealing emergent capabilities like memory, tracking, and compute-aware planning.

◾Reduce 429 errors on Vertex AI: Reduce 429 errors on Vertex AI by choosing the right consumption model, smoothing traffic, and optimizing requests with retries, caching, and global routing for resilience. This post dives into PayGo vs Priority vs Provisioned Throughput, hybrid architectures, and five key techniques like exponential backoff, prompt optimization, and traffic shaping to build scalable, reliable LLM applications.

◾A Coding Implementation to Design Self-Evolving Skill Engine with OpenSpace for Skill Learning, Token Efficiency, and Collective Intelligence: OpenSpace turns AI agents into self-improving systems that reuse skills, cut token costs, and get smarter with every task. This post walks through a hands-on implementation of its evolving skill engine, covering cold and warm starts, custom skill creation, SQLite inspection, cloud-based skill sharing, multi-task pipelines, and benchmark results showing 4.2x income gains and 46% lower token usage.

◾Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled: Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled boosts local coding-agent reliability with better autonomy, stable tool calling, preserved reasoning mode, and native support for the developer role. This post explains how the model was fine-tuned on Qwen3.5 using Opus-style reasoning distillation, what datasets and training setup shaped it, and why it performs well for coding, math, and logic-heavy offline workflows.

◾Join the Gemini Live Agent Challenge: Build immersive multimodal AI agents with Gemini to compete for $80,000 in prizes, Google Cloud credits, and a chance to present at Google Cloud Next ’26. This post announces the Gemini Live Agent Challenge, outlining three categories, prize tiers, and project requirements, while encouraging builders to use Gemini models, the Gen AI SDK or ADK, and Google Cloud services.

◾Deploy SageMaker AI iference endpoints with set GPU capacity using training plans: Guarantee reliable GPU capacity for LLM inference on SageMaker by reserving instances upfront with training plans, ensuring predictable performance and cost control. This post explains how to secure p-family GPUs for inference endpoints, covering capacity search, reservation creation, and deployment workflows, helping teams run uninterrupted evaluations and manage endpoints efficiently during fixed time windows.

◾nvidia/Nemotron-Cascade-2-30B-A3B: Nemotron-Cascade-2-30B-A3B delivers strong open reasoning and agentic performance in a compact MoE setup, with thinking and instruct modes, long context, and standout math and coding benchmarks. This post introduces the model’s capabilities, compares its results against major baselines, and walks through deployment, chat templating, tool use, and agentic setup with vLLM and Transformers.

See you next time!

Share DataPro. Build Your Reputation.

You already know the value of sharp, practical data knowledge. Now help grow the community — and earn a place at the heart of it.

Share DataPro with friends and colleagues. Every subscriber who joins through your link moves you closer to something more than discounts — recognition, authorship, a speaker slot, and insider access to the people shaping data today.

From your first public shoutout at 5 referrals to a named VIP advocate at 300+ — your referrals build your identity in the DataPro community.

Start sharing your link and climb the DataPro Ambassador Ladder.

Why People Champion DataPro

DataPro is built for professionals who want real-world data knowledge, not just theory. When you refer someone, you’re not selling them a subscription — you’re giving them a shortcut to what actually matters in data.

Your Byline. Your Audience. Reach the Contributor tier and co-author an edition of DataPro — distributed to thousands of analysts, engineers, and data leaders. Your insights, your name, your reach.

Step Onto the Stage. Top Champions unlock speaker slots at DataPro workshops and events. Not just attending — presenting. In front of the people who matter in data.

Recognition From Day One Your first 5 referrals earn you a public shoutout in the DataPro newsletter. Not after 100 referrals. Immediately.

Insider Access, Not Just Content Join an invite-only Discord community of analysts, engineers, researchers, and data leaders discussing real challenges and solutions.

Conference & Workshop Access Free tickets to DataPro workshops and events. Early registration and priority pricing before public release.

Expert-Led Content Learn directly from practitioners working on real data systems, pipelines, analytics, and AI.

A Library Worth Owning Packt ebook credits spanning AI, ML, data engineering, and analytics. Permanent resources, not a temporary perk.

Climb the DataPro Ambassador Ladder

Each milestone unlocks a new reward. Think of it as a journey — every referral deepens your place in the DataPro community.

5 Referrals — DataPro Insider “I’m part of something real.”

• Public shoutout in the DataPro newsletter

• Access to the DataPro private Discord community

• Exclusive Insider badge on your profile

25 Referrals — DataPro Contributor “My voice reaches thousands.”

• Co-author a DataPro newsletter edition — your byline, your ideas, your audience

• Free ticket to a DataPro workshop or event

• 50% off DataPro Premium subscription

75 Referrals — DataPro Champion ✦ “I shape how this community grows.”

• Speaker slot at a DataPro workshop or event

• 100% free DataPro Premium — full access, forever

• Named Champion across the DataPro community

• Invitations to special sessions and expert roundtables

150 Referrals — DataPro Scholar “I invest in the craft.”

• One free Packt ebook credit — any title across AI, ML, analytics, or data engineering

• Free ticket to an upcoming DataPro conference or flagship event

• Priority placement for future co-authorship and speaker opportunities

300+ Referrals — DataPro VIP “I am DataPro.”

Top contributors receive:

• Named VIP advocate — recognised across all DataPro channels

• Recurring speaker presence at DataPro workshops and events

• Direct interaction with experts inside the private Discord group

• Invitations to exclusive sessions, closed-door roundtables, and special events

• Free DataPro Premium + all prior tier benefits, sustained

How the Referral Program Works

Step 1 Subscribe to DataPro.

Step 2 Grab your personal referral link from your subscriber dashboard.

Step 3 Share it with friends, teams, Slack channels, and communities.

Step 4 Each new subscriber who joins through your link counts as a referral.

Step 5 Unlock rewards as you hit each milestone.

Who Should Share DataPro

Readers who benefit the most from this program typically include:

• Data analysts

• Data engineers

• AI / ML practitioners

• Product and analytics professionals

• Students learning data skills

• Leaders building data teams

• Data-adjacent roles — product managers, consultants, operators who work with data daily

If your network works with data, your referrals will quickly add up.

Build the DataPro Community

The goal of DataPro is simple:

Bring together people who care about practical data knowledge, expert insights, and solving real problems — and recognise the people who help that community grow.

Every referral strengthens what we’re building together.

Share your referral link and start climbing the ladder. Your first reward is just 5 referrals away.

Start Referring

Open your subscriber dashboard to find your personal referral link and start sharing.

Your first milestone is closer than you think — 5 referrals unlocks your first recognition, your Insider badge, and access to the DataPro private community.

Grow the DataPro community. Earn your place at the heart of it.

Welcome to DataPro #167 – From Flexible Data to Production AI, where we explore how modern data and machine learning systems are evolving to handle dynamic data, reasoning models, and real-world deployment.

Most modern applications no longer operate on perfectly structured datasets. Event streams, product attributes, and application telemetry often evolve faster than database schemas. The challenge today is not simply storing this flexible data, but querying and analyzing it efficiently without sacrificing performance or maintainability.

This week, we’re collaborating with Packt expert author Jun Shan, whose book SQL for Data Analytics (4th Edition) explores practical SQL techniques used in modern analytics systems. In this issue, we look at a capability that has quietly become essential across analytics pipelines and data science platforms: JSON processing in PostgreSQL.

JSON allows teams to store dynamic attributes and evolving event payloads without constantly redesigning schemas. But flexibility alone isn’t enough. As Jun Shan explains, PostgreSQL strikes a balance between relational structure and document-style storage. By combining traditional columns with JSONB storage and proper indexing strategies, teams can support evolving application data while maintaining fast, reliable queries at scale.

In the expert-led piece, you’ll see how JSON processing works in PostgreSQL, how event-tracking systems store dynamic payloads, how SQL can query nested JSON fields directly, and why JSON should complement relational modeling rather than replace it. If you work with PostgreSQL in analytics pipelines, this is a design pattern worth understanding.

Alongside that deep dive, this week’s Data Science & ML Research Roundup explores what’s shaping modern machine learning systems and AI infrastructure:

• Bringing Nano Banana 2 to enterprise
• Operationalizing Agentic AI in production systems
• Building custom model providers for Strands Agents on SageMaker
• Fine-tuning LLMs with Oumi and deploying to Amazon Bedrock
• The Qwen3.5-27B Claude-distilled reasoning model
• Lightricks’ LTX-2.3 audio-visual foundation model

Together, these developments point to a broader shift across the data and AI stack: systems are becoming more flexible, more multimodal, and increasingly designed to move quickly from experimentation to production.

Let’s dive in.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Struggling to move from ML experiments to production systems?

Join this live PyTorch workshop with Ashish Ranjan Jha (Mar 29, 7–10 PM GMT+5 | 9:30 AM–12:30 PM EDT) and build the engineering patterns teams use to ship real ML systems. Save your seat now.

JSON Processing in PostgreSQL

By Jun Shan

Introduction

Modern applications rarely work with perfectly structured data. Instead, they often receive flexible, changing, or partially defined information. This is where JSON processing becomes essential in databases managed by the PostgreSQL DBMS. PostgreSQL allows you to store and query JSON directly, combining the flexibility of document databases with the reliability of relational systems.

Consider a real-world example: an e-commerce platform storing product information. Each product has standard fields like name and price, but also dynamic attributes such as color, size, technical specifications, or regional variations. These attributes differ between products and change over time. Creating separate relational columns for every possible attribute would be inefficient and difficult to maintain.

JSON allows developers to store flexible attributes in a structured format without constantly modifying the database schema. The author, having worked with systems where JSON processing allowed rapid feature development while maintaining strong query performance and data consistency, will share some concepts and tips in this blog.

The Core Concept

JSON processing in PostgreSQL refers to storing, querying, and manipulating data in JavaScript Object Notation (JSON) format directly within the database. PostgreSQL provides two JSON-related data types: JSON and JSONB. JSON stores the data as text, while JSONB stores it in a binary format optimized for fast querying and indexing. In most cases, JSONB is the better choice for production systems.

JSON processing allows PostgreSQL to treat JSON documents as structured data. Instead of retrieving the entire JSON document and parsing it in the application, PostgreSQL can extract specific fields directly within SQL queries.

For example, consider a JSON document stored in a column:

{

“color”: “red”,

“size”: “large”,

“weight”: 2.5

}

PostgreSQL allows you to query individual values inside this document using built-in operators. This allows filtering, searching, and analyzing JSON data efficiently.

JSON processing matters because it provides flexibility without sacrificing database capabilities. Developers can store evolving or optional attributes without constantly redesigning tables. At the same time, PostgreSQL still supports indexing, filtering, and joining JSON data.

However, many practitioners misuse JSON. One common mistake is treating JSON as a replacement for relational design. JSON is useful for flexible attributes, but core relational data such as IDs, relationships, and frequently queried fields should remain in structured columns. JSON is most effective when used strategically, not as a universal replacement for relational structure. Another mistake is storing JSON but not indexing it properly, which leads to slow queries.

Is your Power BI architecture ready to scale?
Design Fabric and Power BI systems that stay fast, governed, and production-ready.

Receive a full architecture playbook, capacity planning worksheet, semantic modeling framework, governance checklist, plus a free Learn Power BI eBook, templates, case studies, and certification. Lock in your seat

Practical Example

Consider a customer event tracking system. The system records user actions such as login, purchase, and page views. Each event has common fields, but also event-specific details.

A table might look like this:

events

------

event_id

user_id

event_type

event_time

event_data (JSONB)

The event_data column stores flexible information depending on the event type.

Example JSON for a purchase event:

{

“product_id”: 12345,

“price”: 49.99,

“currency”: “USD”,

“payment_method”: “credit_card”

}

Example JSON for a login event:

{

“device”: “mobile”,

“browser”: “Chrome”,

“location”: “New York”

}

This design allows the system to store different types of events without changing the schema.

Now suppose you want to find all purchase events where the payment method was credit card. PostgreSQL allows you to query inside the JSON:

SELECT *

FROM events

WHERE event_data->>’payment_method’ = ‘credit_card’;

The operator ->> extracts the value as text, allowing comparison.

If this query runs frequently, you can create an index:

CREATE INDEX idx_events_payment_method

ON events USING GIN (event_data);

This allows PostgreSQL to search JSON content efficiently.

This approach enables flexible storage while maintaining strong query performance. It also reduces schema complexity and accelerates development.

Common Pitfalls

While JSON processing is powerful, improper use can lead to performance and maintainability problems.

One major mistake is storing everything in JSON instead of using relational columns. JSON should not replace core relational structure. For example, fields like user_id, event_time, and event_type should remain separate columns. These fields are frequently filtered and joined, and relational columns perform better for these operations.

Another common mistake is failing to use JSONB. The JSON type stores raw text, which must be parsed every time it is accessed. JSONB stores parsed binary data, which allows faster queries and indexing. Most production systems should use JSONB.

Indexing is another critical area. Without indexes, PostgreSQL must scan every row to find matching JSON values. This becomes slow as tables grow. Using GIN (Generalized Inverted Index) indexes significantly improves JSON query performance.

Large JSON documents also introduce update costs. PostgreSQL stores rows as complete units. When a JSON field is updated, PostgreSQL often rewrites the entire row. Frequent updates to large JSON documents can increase storage usage and reduce performance.

Another pitfall is lack of data validation. Relational columns enforce types such as integer, date, or numeric. JSON fields are more flexible, which can lead to inconsistent data if validation is not enforced at the application or database level.

Experienced practitioners should also watch for query inefficiencies. Extracting JSON fields repeatedly in queries can add overhead. In some cases, frequently queried JSON attributes should be promoted to relational columns.

Effective optimization strategies include:

• Use JSONB instead of JSON

• Index JSON fields using GIN indexes

• Keep core relational fields outside JSON

• Store only flexible or optional attributes in JSON

• Monitor query performance using EXPLAIN ANALYZE

• Avoid excessively large JSON documents

The key is balancing flexibility and performance.

Conclusion

JSON processing in PostgreSQL provides a powerful way to handle flexible, evolving data without sacrificing database performance. It allows developers to store structured information without constantly modifying table schemas, making it ideal for modern applications with dynamic requirements.

The most important insight is that JSON complements relational design rather than replacing it. Core relational data should remain in structured columns, while flexible attributes can be stored in JSONB. When used correctly, PostgreSQL allows efficient querying, indexing, and analysis of JSON data.

The broader impact is significant. JSON processing enables faster development, simpler schema evolution, and better support for modern application patterns such as event tracking, configuration storage, and flexible product attributes. It allows PostgreSQL to combine the strengths of relational and document databases in a single system.

If you work with PostgreSQL, you should actively evaluate where JSON processing can improve your design. Start by identifying flexible or changing attributes, store them in JSONB, and add proper indexes. With careful use, JSON processing can greatly improve both development agility and system performance.

For further information, especially the details of how JSON documents are accessed and transformed, you can refer to Chapter 11, Processing JSON and Arrays in my book, SQL for Data Analytics, 4th Edition, where we provide comprehensive coverage to PostgreSQL’s JSON functionalities.

⚡ Refactor production code safely with AI in our live Packt workshop with ast-grep creator Herrington Darkholme — join us live.

Data Science & ML Research Roundup

◾Bringing Nano Banana 2 to enterprise: Nano Banana 2 promises something image models often miss: understanding the prompt before generating the image. Designed for the new era of generative creativity, it combines Pro-level image generation and editing with the speed of Flash, making advanced visual creation more accessible. This post explores Nano Banana 2’s capabilities, from real-time web-informed visuals and 4K upscaling to consistent subject rendering and accurate text generation, and explains how businesses and creative teams can integrate the model into their workflows to produce high-quality, production-ready visuals at scale.

◾Operationalizing Agentic AI Part 1: A Stakeholder’s Guide. Agentic AI promises transformative productivity, but many enterprise pilots stall before reaching production. This post argues that the gap is not technology but execution. Drawing on AWS’s work with over 1,000 organizations, it explains how leaders can operationalize agentic AI by defining workflows clearly, bounding agent autonomy, measuring outcomes, and treating agents like structured roles within a well-run team.

◾Building custom model provider for Strands Agents with LLMs hosted on SageMaker AI endpoints: Running custom LLMs on SageMaker offers flexibility and cost control, but response format mismatches can prevent them from working with Strands agents. This guide shows how to bridge that gap by building custom model parsers that translate OpenAI-style responses into the Bedrock Messages API format, enabling seamless integration between SageMaker-hosted models and the Strands Agents SDK.

◾Accelerate custom LLM deployment: Fine-tune with Oumi and deploy to Amazon Bedrock. Fine-tuning open-source LLMs often breaks down between experimentation and production due to fragmented tooling. This guide shows how to streamline the process by using Oumi to fine-tune models like Llama on EC2, store artifacts in S3, and deploy them to Amazon Bedrock with Custom Model Import, enabling scalable, managed inference without managing infrastructure.

◾Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled: Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled is a community fine-tuned reasoning model built on Qwen3.5, distilled from Claude 4.6 Opus reasoning traces. This post outlines its training approach, datasets, and improvements in structured chain-of-thought reasoning, autonomy, and coding-agent performance, enabling more efficient step-by-step problem solving in local AI workflows.

◾Lightricks/LTX-2.3: LTX-2.3 is an open audio-visual foundation model from Lightricks that generates synchronized video and sound within a single diffusion-based system. This release improves prompt adherence, audio quality, and visual fidelity, while remaining open-weight and locally runnable, enabling developers to build text-to-video, image-to-video, and multimodal generation workflows.

See you next time!

Welcome to DataPro #166 👋

AI is no longer stuck in the lab. The real challenge now is building systems that scale, stay reliable, and earn trust in production.

This week, we lead with an expert walkthrough featuring Dr. Chanin Nantasenamat, the Data Professor, on why rapid prototyping beats perfection. Using Streamlit, Snowflake, and GenAI, he breaks down how to move from idea to MVP faster, validate with users early, and design AI apps that are production-ready from day one.

If you are building AI agents, fine-tuning models, scaling infrastructure, or optimizing GPUs, this issue connects the dots from front-end experimentation to back-end reliability.

🔎 This Week’s Highlights

• Production-ready AI agents: Lifecycle frameworks for memory, orchestration, evaluation, and deployment.

• Multi-LoRA with vLLM: Efficiently serve fine-tuned MoE models on shared GPUs via SageMaker and Bedrock.

• Claude Opus 4.6 & Sonnet 4.6: Frontier models on Vertex AI for enterprise workflows and agentic systems.

• Amazon Bedrock AgentCore: Secure, scalable, memory-aware AI assistants without custom infrastructure.

• Fine-tuning on Google Cloud: Vertex AI for managed Gemini tuning, GKE for customizable LoRA pipelines.

• CodeFu-7B with RL on SageMaker: Distributed reinforcement learning using Ray and veRL for scalable code models.

• Qwen3.5-397B: A multimodal MoE foundation model with 201-language support and large-scale RL training.

• Meta’s GCM toolkit: Proactive GPU cluster monitoring to eliminate silent hardware failures.

• NVIDIA DreamDojo: A robot world model trained on 44k hours of human video for physics-aware simulation.

Join us this week on February 28th for: GenAI System Design Workshop
Design end-to-end GenAI systems built for scale, cost control, and production accountability.

Walk away with a reusable AI System Design Playbook, real-world architecture case studies, practical decision templates, and direct private chat access to Sairam to clear your doubts. Includes a certificate of completion and post-event resources you can apply immediately. 35% off with code DATAPRO35 | Only 10 seats left.

Sponsor Spotlight

Unblocked: The context layer your AI tools are missing.
Give your agents the understanding they need to generate reliable code, reviews, and answers. Unblocked builds context from your team’s code, PR history, documentation, planning tools, and runtime signals, so AI outputs reflect how your system actually works. See how it works.

All of it points to one thing: AI is becoming a systems discipline. Let’s step into what that looks like in practice.

Cheers,

Merlyn Shelley,

Growth Lead, Packt.

Share Packt DataPro

Dr. Chanin Nantasenamat Explains Why Rapid Prototyping Beats Perfection

In the fifth episode of Pack Talks, host Abhishek Kaushik sits down with Dr. Chanin Nantasenamat, better known to many as the “Data Professor.” With more than 15 years of experience in applied data science, over 160 research publications, and a strong presence in the AI education space through DeepLearning.AI, Snowflake, and YouTube, Dr. Chanin brings both academic depth and hands-on builder energy to the conversation.

The focus of the episode is simple but powerful: how to turn ideas into working AI-powered applications quickly using Streamlit.

What follows is not just a discussion about a Python library. It is a blueprint for how modern builders can move from concept to prototype to MVP in record time.

What Is Streamlit, Really?

At its core, Streamlit is a low-code Python web framework that allows you to build interactive web applications with remarkably little code.

You can create a web app interface with something as simple as:

import streamlit as st
st.title(”My First App”)

That is not an exaggeration. With a few intuitive commands such as st.title(), st.write(), or st.button(), you can generate a clean, modern user interface almost instantly.

What makes Streamlit stand out?

• It is Pythonic. If you know basic Python, you can build apps.

• It is opinionated in a good way. The default UI looks polished without design effort.

• It is open source and free.

• It is easy to deploy and share.

Compared to alternatives like Gradio, Typer, or R’s Shiny, Streamlit strikes a balance between simplicity and flexibility. Shiny caters strongly to the R ecosystem. Gradio is excellent for ML demos. Streamlit, however, feels like a natural extension of Python itself. It is designed for rapid experimentation and iteration.

It is especially appealing to what Dr. Chanin calls “vibe coders” — developers who build interactively, testing ideas in real time rather than engineering everything upfront.

Production PyTorch Workshop
Move from model experimentation to optimized, deployable ML systems with proven export and inference patterns.

Get the complete code repository, deployment checklist, optimization templates, and a FREE Mastering PyTorch eBook.
Includes workshop recording, failure-mode guide, and certificate of completion.
40% off Early Bird with code EBPY40 | Ends next week.

Why Rapid Prototyping Beats Perfection

One of the most important themes in the episode is the value of rapid prototyping.

Many aspiring builders fall into what is often called analysis paralysis. They overthink architecture, scalability, performance, and edge cases before they have even validated the core idea.

Rapid prototyping flips that approach.

Instead of building the perfect system from day one, you:

1. Start with an idea.

2. Build a minimum viable product (MVP).

3. Share it with users.

4. Collect feedback.

5. Iterate.

This cycle dramatically reduces risk.

Rather than relying on your own assumptions, you gather feedback from real users. If 100 people interact with your prototype, their behavior tells you far more than months of isolated planning ever could. You can pivot early. You can refine features that matter. You can drop what does not.

In a world where AI tools accelerate development, the cost of building early prototypes has dropped significantly. That changes the innovation equation.

Making ideas tangible early accelerates product–market fit.

Inside the Fast Prototyping Course

Dr. Chanin’s Fast Prototyping course, available on Coursera and DeepLearning.AI, is structured around this mindset.

The course is beginner-friendly and organized into three main modules:

1. Building an MVP

Learners start by converting abstract ideas into functional web apps. The emphasis is not on theoretical perfection but on building something usable.

2. Iterative Development

The course reinforces the loop: idea → prototype → feedback → improvement.

This teaches learners to think experimentally. A prototype is not a final product. It is a learning instrument.

3. Integration with Snowflake and GenAI

The course then levels up by integrating Streamlit with Snowflake’s AI and data cloud platform. Learners work with:

• Databases

• APIs

• AI inference using Snowflake Cortex

• Real-world datasets such as an Avalanche winter gear dataset

The progression is deliberate. First, build something. Then make it intelligent.

Importantly, the course does not assume deep knowledge of data structures, algorithms, or advanced machine learning. Basic Python familiarity and some exposure to libraries like Pandas and NumPy are sufficient.

There is also a complete code repository, allowing learners to run demos, experiment, and adapt examples without starting from scratch.

Streamlit as the Front-End Layer of AI Systems

One of the most valuable conceptual takeaways from the episode is architectural clarity.

Streamlit is not the AI model. It is not the database. It is not the inference engine.

It is the user interface layer.

Behind the scenes, your application might include:

• Large language models for inference

• Databases for structured data

• APIs for external services

• Retrieval-augmented generation (RAG) pipelines

• SQL generation from natural language queries

Streamlit orchestrates the interaction between user and system.

For example, in the course, learners build a chatbot that converts natural language questions into SQL queries. A user asks a question. The model translates it into SQL. The database retrieves structured answers. The results are displayed instantly in the Streamlit interface.

This bridges the gap between complex backend AI workflows and intuitive user experiences.

Power BI & Fabric Architecture Workshop
Architect scalable, governance-ready BI systems with the right Fabric, capacity, and semantic modeling decisions.

Receive a production-ready architecture playbook, capacity planning worksheet, semantic modeling framework, and governance checklist.
Includes FREE Learn Power BI eBook, templates, case studies, and certificate of completion.
40% off Early Bird with code EARLY40 | Limited time.

Prompt Engineering and Observability

As GenAI systems become more integrated into applications, prompt engineering becomes critical.

How you structure instructions to large language models affects:

• Output quality

• Accuracy

• Reliability

• User trust

But there is another crucial layer: observability.

When building AI systems, you need to understand:

• What inputs were given to the model?

• What outputs were generated?

• Which data sources were used?

• How consistent are responses?

Tools such as TruEra, an open-source observability platform acquired by Snowflake, help track model behavior and inputs and outputs. Observability strengthens trust and makes debugging easier.

When you combine:

• Streamlit as the UI layer

• Prompt engineering for better model outputs

• Observability tools for transparency

You create AI-powered apps that are not only functional but trustworthy.

Reading along? Catch the full article on our Packt Medium handle.

Data Science & ML Research Roundup

⚫ A dev’s guide to production-ready AI agents: AI agents have evolved from research prototypes to real-world products, raising new challenges in testing, memory, orchestration, and security. Unlike deterministic software, agents require lifecycle-specific frameworks. This guide collection covers architecture, interoperability, context engineering, evaluation, and production deployment, helping developers move confidently from prototype to scalable, trustworthy agent systems.

⚫ Efficiently serve dozens of fine-tuned models with vLLM on Amazon SageMaker AI and Amazon Bedrock. Running multiple fine-tuned MoE models can waste GPU capacity when traffic is low. New multi-LoRA support in vLLM enables several customized models to share a single GPU by dynamically swapping lightweight adapters. With kernel and execution optimizations, this approach improves latency and throughput, making large-scale, cost-efficient AI deployment far more practical today.

⚫ Expanding Vertex AI with Claude Opus 4.6. Google Cloud has added Anthropic’s Claude Opus 4.6 and Sonnet 4.6 to Vertex AI, expanding its frontier model lineup. Opus targets complex enterprise workflows and agentic systems, while Sonnet balances speed and cost. With managed infrastructure, governance, long context windows, and agent tooling, Vertex AI enables secure, scalable production AI across coding, analysis, and automation use cases.

⚫ Building intelligent event agents using Amazon Bedrock AgentCore and Amazon Bedrock Knowledge Bases. This post shows how to build and deploy a production-ready, personalized event assistant using Amazon Bedrock AgentCore. Instead of spending months on infrastructure, teams can use managed memory, identity, runtime, and RAG components to deliver secure, scalable AI agents. The solution supports thousands of concurrent users with enterprise-grade security, session isolation, and long-term personalization.

⚫ Mastering Model Adaptation: A Guide to Fine-Tuning on Google Cloud. As AI apps move to production, prompt engineering often falls short on consistency and control. Google Cloud’s new labs show how to fine-tune models for reliability and specialization. Choose Vertex AI for a fully managed Gemini tuning workflow, or GKE for customizable LoRA-based tuning of open-source models, enabling production-ready, domain-adapted AI systems.

⚫ Train CodeFu-7B with veRL and Ray on Amazon SageMaker Training jobs. Training advanced code-generation models with reinforcement learning is complex and infrastructure-heavy. This post shows how to train the 7B CodeFu model using veRL and Ray on Amazon SageMaker, combining distributed RL orchestration with managed infrastructure. The result is scalable, fault-tolerant training for algorithmic reasoning models, without managing clusters manually.

⚫ Qwen/Qwen3.5-397B-A17B: Qwen3.5 marks a major upgrade in foundation models, combining multimodal early fusion, a hybrid Gated DeltaNet plus sparse MoE architecture, large-scale reinforcement learning, and support for 201 languages. With 397B parameters and efficient activation, it delivers strong reasoning, coding, vision, and agent performance while optimizing throughput, latency, and global accessibility.

⚫ Meta AI Open Sources GCM for Better GPU Cluster Monitoring to Ensure High Performance AI Training and Hardware Reliability: As AI training clusters scale to thousands of GPUs, hardware instability becomes a hidden risk. Meta’s open-source GCM toolkit tackles “silent” GPU failures by tightly integrating hardware telemetry with Slurm-based HPC orchestration. With automated health checks, DCGM diagnostics, and OpenTelemetry support, GCM enables reliable, large-scale model training and better infrastructure observability.

⚫ NVIDIA’s DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos. DreamDojo is a large-scale robot world model trained on 44,000 hours of egocentric human video to simulate dexterous, contact-rich tasks. Using continuous latent actions and a real-time distillation pipeline, it enables physics-aware, controllable predictions for robotics. This advances scalable robot training, teleoperation, and model-based planning in open-world environments.

See you next time!

As AI generates more learning content, it is becoming harder to see where expert input really makes a difference. Packt has recently partnered with Go1 to create a short study looking at how developers actually learn today, and when structured courses still matter alongside AI tools.

If you work with learning or rely on it to build skills, your perspective would be useful. The survey takes under 5 minutes to complete, and the results will be shared in a study published in March.

Take the 5-Minute Survey

Subscribe | Submit a tip | Advertise with Us

👋 Hello,

Welcome to DataPro #164. As data science and ML roles continue to evolve, standing out increasingly depends on how well you can demonstrate real-world problem solving, not just model knowledge.

So, if you’ve ever wondered whether Kaggle is “just competitions,” or why so many strong data scientists and ML engineers still credit it for major career breakthroughs, this issue is for you.

In this edition, Luca Massaron, co-author of The Kaggle Book, 2nd Edition, breaks down what Kaggle truly offers beyond leaderboards. He explains how notebooks, datasets, and competition workflows help build a visible record of problem solving, experimentation, and technical judgment. More importantly, he shows how this experience translates into real-world skills and interview-ready stories using the STAR framework.

To support your learning, our authors have also created a free reference cheatsheet that maps all the libraries covered in the book, giving you a clear learning path as you work through the resources. You can download it here: Kaggle Book Cheatsheet.

💡 Workshop Spotlight: Machine Learning and Generative AI System Design Workshop

Join Sairam Sundaresan, AI Engineering Leader, for a hands-on system design workshop on February 28, focused on building machine learning and generative AI systems that actually scale. In this live session, you’ll move beyond model demos and learn how experienced architects design end-to-end AI systems by balancing cost, latency, quality, and risk.

Through guided exercises and design sprints, you’ll practice making real architectural trade-offs and defining success metrics that go beyond accuracy. Whether you’re an ML engineer, architect, or product leader, this workshop equips you with reusable frameworks to design AI systems that hold up in production and evolve with changing models and regulations.

Register Now and Save 35%

Use DATAPRO35 at checkout for early access savings and reserve your seat.

Cheers,

Merlyn Shelley

Growth Lead, Packt

The Essential Asset: Leveraging Kaggle Experience in a Competitive Professional Landscape

Engaging with the Kaggle platform offers a clear advantage for data science professionals seeking to emerge in complex and challenging job market situations, such as the recent one marked by widespread layoffs and hiring difficulties. While competitive data science does not cover the entire span of enterprise-level processes related to data processing and MLOPs, the knowledge and skills acquired on Kaggle are a significant complement to real-world experience. Kaggle serves as an integrated environment for acquiring, documenting, validating, and showcasing competencies that help candidates stand out from the crowd and avoid becoming obsolete in front of automated machine learning (AutoML) or other off-the-shelf solutions, such as recent tabular AI solutions.

The Creation of a Verifiable Portfolio

Employers often view a robust portfolio of projects as a great demonstration of technical knowledge and hands-on experience compared to academic credentials alone. Kaggle facilitates the building up of this critical professional asset.

First of all, Notebooks are recognized as the most important tool (after rankings) for demonstrating a candidate’s abilities, providing tangible evidence of their capacity for clean coding and effective communication. Even if top ranks are not achieved, high-quality Notebooks focused on Exploratory Data Analysis (EDA), tutorials on model architectures, or implementations of cutting-edge research prove crucial abilities, such as extracting visual and non-visual insights from data. Notebooks showcase not just what a candidate has done, but also how they approach problems and communicate insights and conclusions, which is critical for working with management, clients, and experts from diverse backgrounds in a business-oriented company environment.

On the other hand, Kaggle Datasets provide an excellent means for demonstrating ability with data to be used with Machine Learning (ML) algorithms. By curating, cleaning, and documenting data, professionals can publish and maintain a dataset on Kaggle, thereby demonstrating their understanding of the data’s value and potential. The presence of a description, tags, a license, sources, and a frequency of updates are pieces of information used to calculate a usability index, which helps others understand how to use the data. This shows an ability to manage and document data over time. Recently, this opportunity has also been extended to models, allowing the showcase of the necessary competencies in maintenance, fine-tuning, and evaluation of both small and large language models.

This Week’s Sponsor: Progress Telerik

Webinar: How to Build Faster with AI Agents
Learn how full-stack developers boost productivity by up to 50% using AI agents to automate layout, styling, and component generation with RAG and LLM pipelines.
See how orchestration and spec-driven workflows keep quality and consistency in check. Save your seat.

Accelerated Skill Acquisition and Marketability

Kaggle participation fosters self-development, exposing data scientists to diverse data types and problems, demanding rapid iteration on model hypotheses, and requiring extensive feature engineering, experience akin to “competition heat.” This challenging environment sharpens skills necessary for finding quick and effective solutions to data problems.

For job seekers, this translates directly into marketability. Recruiters and human resource departments often monitor Kaggle profiles and rankings when searching for candidates with specific or rare competencies, such as those demonstrated in NLP or computer vision competitions. Consistently good performance in multiple competitions signals a genuine competency and provides verifiable credentials that differentiate an applicant from the crowd.

Furthermore, teaming up in competitions teaches individuals to work collaboratively toward a common goal within a limited time frame, and teamwork is a highly valued quality in data science teams. Participating in Kaggle competitions also enhances networking opportunities, facilitating connections that may result in job referrals and opportunities.

The Kaggle Book, 2nd Edition isn’t just about winning competitions. For data scientists and ML engineers, it offers a practical way to deepen modeling intuition, experiment with real world datasets, and refine end to end problem solving through notebooks and iterative workflows.

As data science roles increasingly demand production awareness, rapid experimentation, and clear communication of results, this book helps you build skills that translate directly into stronger models and better technical decisions. With 30% off the eBook and 20% off the print edition, it’s a timely opportunity to add a structured, hands on reference to your learning stack.

Add to Cart

Translating Experience into Interview Gold via the STAR Approach

The experience gained on Kaggle is invaluable during the job interview process. Candidates should leverage their competition efforts to demonstrate problem-solving capabilities using the STAR (Situation, Task, Action, Result) approach. This approach requires structuring competition narratives to emphasize past behavior rather than simply reciting technical capabilities.

For example, when detailing a challenging competition:

Situation: The candidate must provide a clear context for the problem encountered, detailing the environment and why the situation required attention or action.

Task: Clearly explain the objective taken on, such as cleaning messy data, doing explorative analysis (EDA), or continuously improving a benchmark model.

Action: Describe the specific steps executed. This can involve explaining the methodologies or media (such as notebooks) used to implement the solution.

Result: Articulate the achievement, whether it was improving business value, beating a reference benchmark, or learning from the challenges faced.

By utilizing the STAR framework with Kaggle examples, professionals can craft compelling narratives that effectively articulate their problem-solving capabilities and capacity for incremental improvements, thus granting them an edge over other applicants in a very competitive hiring landscape.

Wrapping up how Kaggle can give a boost to your career

The ability to build a robust portfolio, rapidly acquire new skills, and articulate experiences with clarity and confidence are timeless assets in any competitive field. Kaggle provides a unique and effective arena for developing these very competencies. The platform’s emphasis on tangible results and peer-reviewed work ensures that the skills showcased are not merely theoretical but demonstrably real. For professionals committed to lifelong learning and staying ahead of the curve, engaging with Kaggle is a direct investment in their career longevity and relevance. By translating this experience into compelling narratives, as outlined through the STAR approach, candidates can effectively communicate their value, demonstrating that they are not just spectators in the data science field but proactive actors in its evolving future.

Reading along? Don’t forget to download the free Kaggle Book Cheatsheet a clear reference point for the libraries covered throughout the book.

See you next time!

The hardest part of building GenAI systems is not getting them to work.
It is getting them to keep working after launch.

Many teams ship GenAI applications that perform well in controlled tests and early demos, only to struggle once real users arrive. Latency increases, costs spike, outputs become inconsistent, and small changes have outsized effects. What looked stable during prototyping becomes fragile in production.

This gap between demos and durable systems is exactly what Packt’s Machine Learning & Generative AI System Design Workshop is designed to address. Led by Sairam Sundaresan, an AI Engineering Leader with hands-on experience designing and scaling production AI systems, the workshop focuses on how to move beyond model-centric thinking and design GenAI systems that actually work in the real world.

If your GenAI work needs to survive production, this is your next step. Book now and save 35% with code FLASH35.

A production GenAI application is not a single model behind an API. It is a system composed of retrieval pipelines, embeddings, prompts, infrastructure, evaluation logic, and cost controls. Each component interacts with the others. Small design decisions compound quickly, which is why teams using the same model often experience very different outcomes in production.

Now, let’s look at an example.

A team builds a GenAI assistant for internal documentation and support queries. In early testing, it works well. The model answers questions accurately, and the demo lands perfectly with stakeholders.

To improve reliability, the team adds Retrieval Augmented Generation. Responses become more grounded. Confidence increases. The system ships.

Then real usage begins.

Latency creeps up because every request now includes retrieval and reranking. Some queries pull irrelevant context, which confuses the model instead of helping it. Updating documentation means re-embedding large portions of data. Costs scale linearly with usage. When answers are wrong, it is unclear whether the issue lies in retrieval, prompting, or the model itself.

Nothing has failed outright, but the system becomes fragile. Small changes produce unpredictable behavior. Debugging turns into guesswork.

This is not a tooling problem. It is the result of adding RAG without designing the system around it. RAG is not a feature. It is an architectural decision that affects latency, cost, evaluation, and long-term maintainability.

This is also why accuracy alone is a weak measure of success in GenAI systems. Accuracy does not capture latency under load, cost at scale, consistency across users, or failure modes. A response can be correct and still be unusable. In production, success is defined by system behavior, not model output.

If your GenAI work needs to survive production, this is your next step. Book now and save 35% with code FLASH35.

Many teams encounter another failure when they upgrade models. Prompts behave differently. Context assumptions break. Evaluation metrics stop reflecting user experience. The system collapses because it was designed for stability in an environment defined by change. Future-proofing GenAI systems requires architectures that expect evolution.

System design skills are rarely taught explicitly. Most practitioners learn them through failure, often after systems break in production. In GenAI, that learning curve is expensive.

In this 4.5-hour, live, hands-on workshop, you will design GenAI systems end to end, reason through trade-offs across cost, latency, retrieval, and evaluation, and explore real-world failure modes before they show up in production. The focus is on building reusable mental models, not following tools or trends.

GenAI systems do not fail because the technology is unreliable. They fail because system design is treated as an afterthought. If you want your GenAI work to survive production, learning to think like a system designer is no longer optional.

If your GenAI work needs to survive production, this is your next step. Book now and save 35% with code FLASH35.

See you at the workshop.