However, what is undeniable is that companies have invested significant time, effort, and money into supporting the growth of these frontier technologies and, subsequently, the infrastructure needed to power them.

I recently joined a program called the Venture Institute by Decile Group, an 8-week, no-cost intensive helping me understand the various areas of venture capital while applying these skills in the real world. One of the sprints had us create a fund thesis, with a specific geography, stage, and niche we’d focus on throughout the program. I chose to focus on U.S. seed-stage AI physical Infrastructure Software for a few reasons.

And yes - one of my first Second Layer Spotlights was centered on AI Infrastructure; however, this week’s segment is a bit more focused: asset-light SaaS companies building the connective-tissue layer between AI compute demand and the energy/grid/cooling systems necessary to sustain them.

Why this specific sub-segment? There are a few reasons:

• More capital-intensive projects mean longer payback periods, higher capital needs from investors, and less sustainable revenue cycles. At the seed stage, I felt there were too many risks involved; furthermore, from a funding standpoint, a seed-stage fund would prefer to make many smaller bets rather than concentrate on a few larger ones.

• It’s easier to generate traction with SaaS companies, and since I’m focusing on seed-stage over pre-seed, evidence of some form of traction (customer contracts, pilots, purchase orders, studies) is necessary for me to consider a firm within the range of my thesis.

• The problem isn’t solely what’s built - but how what’s built is optimized to sustain new forms of efficient, clean energy. For example, 70% - 90% of renewable energy projects fail before they even start (Paces, 2025). Additionally, only 19% of power projects in interconnection queues from 2000 to 2018 were actually built by the end of 2023, reflecting lengthy wait times and the need for optimization. In other words, the energy capacity is there - but a lengthy rollout process causes less than 1 out of 5 of all projects to make it past the finish line.

AI is causing a surge in energy demand: US electricity demand is growing 3x to 4x faster than in previous years, a demand largely increased by a need to power AI/data centers (Tanner et al., 2026). By 2030, 25% of all US energy demand will come from data centers (Hanwha, 2025). The demand is there - the question is, will the supply keep up?

Startups, corporations, and investors can’t solely focus on building these systems: helping data centers and their respective cooling systems remain efficient amidst surging demand is just as essential. Additionally, enabling the swift mobilization of energy projects remains just as crucial. Here are 3 seed-stage startups addressing these problems at their core.

Subscribe now

1. Glacian Technologies

$150,000, Seed. Pittsburgh, PA. Founded by Jie Zhao (CMU PhD, serial entrepreneur, President) and Wangda Zuo (Penn State professor, CTO). Investors include the Tartan Entrepreneurs Fund (Carnegie Mellon). Separately backed by $8M+ in non-dilutive foundational research grants (DOE, NSF, DOD, JPMorgan Chase) — not dilutive venture capital investment.

Cooling systems represent upwards of 40% of a data center’s total electricity consumption. Glacian Technologies helps data centers reduce their cooling power consumption by fusing physics-based thermodynamic models with AI/ML optimization, without adding additional hardware. By referencing climate and economic data (electricity prices and ambient temperature), Glacian Technologies more efficiently adjusts cooling in real time.

Glacian has already been implemented into Alerify, a data center in Harrisburg, Pennsylvania, with a peer-reviewed study demonstrating a 25% reduction in cooling energy use. Glacian Technologies spun out of Penn State College of Engineering and has secured a significant number of non-dilutive research grants.

Some of the top-scoring factors for Glacian Technologies include:

• Founder Market Fit: Dr. Wangda Zuo, CTO and Co-founder, brings 10+ years of domain-specific research in data center cooling and building energy models. Dr. Jie Zhao, a successful serial entrepreneur and Carnegie Mellon PhD, brings 15+ years in commercializing research. Great commercial/research blend.

• Technical Differentiation: Physics-informed reinforcement learning (as opposed to the pure machine-learning focus of competitors) has enabled an impressive 25% reduction in data center cooling energy.

• Capital Efficiency: $150k in angel funding for their demonstrated traction is very impressive. Additionally, $8 million in non-dilutive research grants is very favorable for seed-stage investors.

2. CapeZero

$2.6 million, Seed. New York, NY. Founded by Manish Hebbar (CEO, 15 years in renewable finance, 60+ deals / 11 GW closed) and Sumit Chachra (CTO). Investors include Powerhouse Ventures (lead), Climactic, Avesta Fund, Virta Ventures, Stepchange.

Mapping out the financials of clean energy projects has traditionally proven significantly burdensome, contributing to the high likelihood of these projects failing.

Sparked by a desire to preserve a healthy planet for future generations, Manish Hebbar and Sumit Chachra launched CapeZero to help finance and development teams better understand the feasibility of their clean energy projects, regardless of how early or late they are in the process.

Despite its early stage, CapeZero has already partnered with Third Pillar Solar, a major renewable energy developer funded by Glentra Capital. In just 6 months, Third Pillar Solar has reported significant workflow enhancements as a result of CapeZero’s integration, making it easier for their project managers, investors, and finance team to both evaluate and communicate regarding new infrastructure opportunities.

The three top-scoring criteria from the Second Layer Evaluation:

• Founder Market Fit: Manish Hebbar and Sumit Chachra have managed $16 billion+ across 11 GW of power before founding CapeZero. They have lived the problem, and are creating what they wish they had.

• Product Market Fit: Third Pillar Solar serves as fantastic validation, with a report showing CapeZero’s integration significantly enhanced workflow visibility and accelerated project timelines. 50% - 75% speed improvement on deals is real validation.

• Investor Signal: Powerhouse Ventures is a notable energy investor, with various successes including Pearl Street Technologies (acquired by Enverus) and the scaling of Terabase Energy and Overstory.

3. Aravolta

~$5.1 million, Seed (YC S25). San Francisco, CA. Founded by Margarita Groisman (CEO, ex-Microsoft data center chip/server deployments), Jack Sutton (CTO), and Pietro Sette (COO). Investors include Y Combinator, Pioneer Fund, Susa Ventures, Afore Capital, Banyan, Crucible, Topology, plus Guillermo Rauch (Vercel CEO) and Nick Hansen.

Aravolta provides unified data center software to optimize power and cooling efficiency designed for AI-era GPU-dense facilities. While most data center operators run separate tools for building management, power monitoring, Supervisory Control and Data Acquisition, network operations, and asset tracking, Aravolta provides them all under a single lightweight utility node. This allows operators to view the different components of data center operations under a single roof.

Aravolta has significant evidence of early success across data center integrations, serving as early proof that their technology works. As an API-first platform, Aravolta eliminates the need for per-point pricing through the use of pre-built device templates, enabling the onboarding of building, electrical, and IT assets in as little as a few hours.

Here are the standout factors for Aravolta:

• Founder-Market Fit: CEO Margarita Groisman (female founder) ran chip and server deployments inside Microsoft’s data centers, addressing the pain points she used to experience directly.

• Product-Market Fit: multiple named paying customers (OpenColo, Flexnode, DataCrunch, Centra) demonstrate the product aligns with the needs of data center operators, while providing the team good quality feedback.

• Traction: with SOC 2 Type 1 and Type 2 certifications, Aravolta has proven across its early integrations that not only are they successful at helping teams manage real-time data center operations, but they do so with the highest data integrity concerns in mind.

Closing Thoughts

There’s been a lot of valid concern over the construction of data centers and their impact on consumer cost and the environment. It’s for this reason that I wanted to focus within this space: how can companies help shift the narrative? How can we ensure these systems run efficiently to the benefit of all parties?

The future of AI isn’t just ensuring the demand side is met. It’s about allowing the supply side to operate as effectively as possible. I don’t think there’s enough focus on grid optimization, and I’m excited to see how these three companies change that.

References

Hanwha. (2025, February 25). What's in a number: 5 figures on the next era of AI data centers. Hanwha Newsroom. https://www.hanwha.com/newsroom/news/feature-stories/whats-in-a-number-5-figures-on-the-next-era-of-ai-data-centers.do

Lawrence Berkeley National Laboratory. (2026, July 1). Backlog of power plants seeking transmission grid connection eased somewhat in 2025 amidst high withdrawals. Energy Markets & Planning. https://emp.lbl.gov/news/backlog-power-plants-seeking-transmission-grid-connection-eased-somewhat-2025-amidst

Oort Alonso, A., Baranko, K., & Wagler, T. (2025, April 22). Pre-development at scale: Modeling risk in early-stage solar development. Paces. https://www.paces.com/white-papers/pre-development-at-scale-modeling-risk-in-early-stage-solar-development

Tanner, B., Belle, D., Kerry, C. F., Kyosovska, N., Renda, A., Tabassi, E., & Wyckoff, A. W. (2026, April 10). Global energy demands within the AI regulatory landscape. Brookings Institution. https://www.brookings.edu/articles/global-energy-demands-within-the-ai-regulatory-landscape/

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Website: https://bryan-hanley.lovable.app/

Leave a comment

A June 2026 study evaluated the difference in usage between two of OpenAI’s flagship products over 28 days: ChatGPT, known for conversational/prompt-based answers, and Codex, which focuses on agentic work. Within OpenAI (as a corporation), Codex accounted for 99.8% of OpenAI workers’ output tokens; for organizational users (users on Business or Enterprise plans), this number was 63.3% (Johnston et al., 2026).

That same study found that the number of active Codex users has grown more than fivefold in the first half of 2026 (Johnston et al., 2026). It’s clear organizations are recognizing the value of agentic systems; this number is only expected to increase over the next few months as more users, businesses, and enterprises alike hand off their workload to agents.

However, as the saying goes, “with more power comes more responsibility”. As we continue to allocate more responsibility to AI agents, we must ensure these agents have the tools and infrastructure available to execute both properly and safely.

The following 3 seed-stage companies aren’t trying to outcompete Codex or Claude Code: they’re developing the infrastructure for these frontier technologies to run as they should.

Subscribe now

Nekuda

$5 million, Seed. New York, NY. Investors include Madrona (lead), Amex Ventures, Visa Ventures, Paul Klein (Browserbase CEO), Shyamal Anadkat (OpenAI Applied AI), Sahar Mor

Ayal Karmi (CEO and Co-founder) was experimenting with his own AI shopping agent to order a pizza, and discovered a major problem: checkouts were never meant for AI agents. Every agent that attempts to transact end-to-end simply cannot - and Karmi and his co-founders recognized the greater implications as agents take on more responsibility.

Nekuda provides the payment-infrastructure Software Development Kit that lets AI agents make real online payments on a user’s behalf. It securely stores/injects payment credentials, gets instant user authorization through agentic mandates, and layers on compliance and trust guardrails to ensure transactions are made safely and securely.

Serving as the Stripe/Plaid equivalent for AI-agent-initiated transactions, Nekuda is a first-mover taking advantage of a rapidly expanding market.

Here are the factors that stand out with Nekuda when analyzed under the Second Layer early-stage startup assessment:

• Founder-Market Fit: Ayal Karmi (CEO) spent 2011–2015 in the FX Department at the Central Bank of Israel, then seven years as co-founder/COO of Taranis (agricultural computer vision), then COO/VP Operations at Solidus Labs (crypto market surveillance) from 2021–2023.

• Timing & Competition: Payment rails (Stripe, Visa, Mastercard) and agent frameworks (LangChain, OpenAI Agents SDK) each assume the other will handle agent-authorized transactions, and neither has built it. Nekuda’s bet is that the missing primitive isn’t authentication but authorization — proving what the user pre-approved the agent to spend on — and it needs to be a dedicated SDK before the rails absorb it.

• Traction Quality: Rare strategic-investor combination at seed — both Amex Ventures and Visa Ventures on the cap table alongside Madrona (lead), with angels including Paul Klein (Browserbase CEO), Sahar Mor (ex-Stripe), and Shyamal Anadkat (OpenAI Applied AI). Named launch partner for Visa Intelligent Commerce and wallet infrastructure behind both Visa Intelligent Commerce and Mastercard AgentPay.

Langtrace

$5.3 million, Seed (specific to their original crypto-observability business, Scale3, from which Langtrace branched off). San Francisco, CA. Investors include Redpoint Ventures, Mysten Labs, and Howard University.

When the founders worked at Coinbase, they realized there was no solid tooling for observing blockchain nodes at scale. As LLMs scaled, they realized an identical pattern: every time LLMs moved from a development to a production environment, their deployment was plagued with issues.

The founders bet that LLMs need a new observability tool built on the same open standard as enterprise infrastructure on which it runs, rather than an external tool. They once again saw this problem develop amongst agentic workflows. Langtrace traces every step an agent takes: LLM calls with prompts, completions, tokens, tool calls, augments passed, and more.

Langtrace provides users and corporations with the “flight logs” of LLM and agentic workflows: what went wrong? Where did the agent hallucinate? And what part of the process was broken? Understanding these components makes it easier to address and prevent future issues.

• Founder-Market Fit: Karthik Kalyanaraman (CTO) was Technical Lead on Coinbase’s Infrastructure/Observability team, building blockchain-node observability at production scale (prior: HP, VMware). Ola Muse (CEO) was on the same Coinbase team - teammates from former employers correlate relatively highly with startup success.

• Timing & Competition: the agentic-workflow wedge - Single-LLM observability is now crowded (LangSmith, Arize, Langfuse, Braintrust, Helicone all $10M+), but agent observability is materially harder — failures hide in intermediate tool calls, retrievals, and subagent handoffs, not the final answer.

• Traction Quality: MongoDB engineering publicly documents Langtrace as a reference AI-observability integration (matters because Atlas Vector Search sits in many agentic stacks). That point of reference is highly valuable, especially at the seed stage.

AgentMail

$6.0 million, Seed. San Francisco, CA. Investors include General Catalyst, Y Combinator, Phosphor Capital, Paul Graham, Dharmesh Shah, Paul Copplestone, and Karim Atiyeh.

As the founders were building early agent prototypes, they realized the consistent need for human input. Integrating agents into human workspaces proved challenging - put simply, agents were never taught how to communicate with humans. Currently, methods of doing so are clunky, complicated, or overcrowded.

So, Haakam Aujla and his team built dedicated email inboxes for AI agents. This includes full two-way email with parsing, threading, labeling, search, and replying. The tech stack is built as an API so any agent framework can create a new communication system as opposed to hijacking a shared human inbox.

With a very strong VC/angel backing for added credibility, AgentMail is tackling an emerging problem that is going to proliferate as agents are asked to take on more and more complex responsibilities.

• Founder-Market Fit: Trio combines Accel/StepStone investing pattern-matching (Singh), NVIDIA autonomous-vehicles + Meta engineering (Kim), and Optiver quant-research rigor (Aujla) — a rare mix of investor instinct, hard technical infra, and quantitative depth for a seed team. University of Michigan cohort.

• Timing & Competition: Every agent needs a real communications channel to loop humans back in, currently duct-taped through shared inboxes or transactional-email APIs that assume one-way sending. No email incumbent (Gmail, Outlook, Superhuman) is building agent-first inboxes. AgentMail’s wedge: own the abstraction now, while the pattern is still forming and before Postmark/Sendgrid retrofit it.

• Traction Quality: 500+ B2B customers, tens of thousands of humans, and hundreds of thousands of agent users within ~6 months of launch. Users tripled the week OpenClaw went viral (late January 2026) and quadrupled the following month — direct correlation to the agent-adoption curve. Strategic angel roster is category-defining: General Catalyst (lead), Y Combinator, Paul Graham, Dharmesh Shah (HubSpot CTO), Karim Atiyeh (Ramp CTO), Paul Copplestone (Supabase CEO), Taro Fukuyama.

Closing Thoughts

As agentic workflows continue to scale, consumers must consider the infrastructure layer behind them. The above companies may not be the household names of OpenAI or Claude, but without them, agentic workflows will never reach their full potential. Investors should prioritize the backend just as much as the front. I’m excited to watch these companies scale in accordance with the growing integration of agentic workflows.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

Reference

Johnston, D., Holtz, D., Richmond, A. M., Ong, C., Tambe, P., & Chatterji, A. (2026). The shift to agentic AI: Evidence from Codex (arXiv:2606.26959). arXiv. https://arxiv.org/abs/2606.26959

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Website: https://bryan-hanley.lovable.app/

Leave a comment

But what really stuck with me about that conversation was how it ended — he said, “I wonder how long that’ll last.” It’s something I hadn’t really thought about before then, but seeing the valuations of companies like OpenAI, it’s certainly gotten my attention now.

Subscribe now

Seismic Valuations

As of late March 2026, OpenAI secured $122 billion in committed capital at a valuation of $852 billion — the largest private funding round in AI history (OpenAI, 2026). Make no mistake, OpenAI is a revolutionary technology, and it’s clear why the company has become a household name.

However, we need to look at this from the investor’s perspective too. Investors expect a path to profitability and a return on that capital. For OpenAI, that path is a long one: the company is projected to lose $14 billion in 2026, with cumulative losses expected to hit $44 billion between 2023 and 2028. Profitability, at the earliest, isn’t expected until 2029 (Dubois, 2026).

Sticking with the OpenAI example, only 5.5% of ChatGPT’s 900 million weekly users pay for a subscription, leaving OpenAI to bear the compute cost of serving the remaining 94.5% for free (Dubois, 2026). Something will eventually have to change, or investors will look elsewhere. That change could take the form of higher consumer prices, stricter usage limits, or new paywalls altogether. (Note: I’ve seen references to OpenAI introducing ads, but I wasn’t able to confirm the details or a solid source for this — worth verifying before you publish, or cutting the line if you can’t nail it down.)

Could AI Be More Expensive than Humans?

Enterprise-focused business models have delivered significantly better returns for AI companies than consumer-focused models have. But for many organizations on the buying side, the cost of compute has turned out to be steps beyond the cost of payroll… and for many companies, that wasn’t expected.

Take Uber, for example: the company’s CTO burned through its entire 2026 AI coding budget in just four months. With Claude Code adoption rising sharply across its 5,000-engineer team, and per-engineer monthly API costs climbing as high as $2,000, the accelerated spend upended the company’s budget entirely (Investing.com, 2026).

For many corporations, AI still functions more like a corporate talking point than a significant value-add. MIT’s NANDA Initiative found that 95% of enterprise generative AI pilots deliver zero measurable return on profit and loss (Field, 2025).

While there’s certainly a learning curve to integrating AI into existing workflows, the productivity gains touted by corporate leaders so far reflect a Pareto distribution more than an “everyone wins” story — most organizations simply aren’t seeing a significant impact yet.

A Changed Job Narrative

If you remember, just a few months ago the AI “jobpocalypse” was a major news story, with corporate leaders warning that AI would eliminate millions of jobs across the United States, particularly for younger graduates. It turns out that reality hasn’t unfolded — at least not yet.

Apollo Global Management chief economist Torsten Sløk has said there’s “zero evidence” AI is killing jobs (Fortune, 2026b). In fact, the data suggests a different reality, in which jobs are being created rather than destroyed.

Anthropic’s own research found no systematic increase in unemployment among highly exposed workers since late 2022, and PwC’s 2025 Global AI Jobs Barometer, which analyzed nearly a billion job ads, concluded that AI is actually making workers more valuable, not less (”Dream Machine,” 2026).

That said, the picture isn’t uniformly rosy. Some younger workers in the most AI-exposed occupations did see a decline between November 2022 and December 2025, but this appears to be driven by fewer job openings for new entrants rather than layoffs of existing employees (”GAGI,” 2026). For younger workers and recent or future graduates, learning to work alongside these tools remains crucial.

Moving Forward

There’s so much still to uncover about AI’s current and future impact on the workforce, and on the investors funding its growth. In this article, I wanted to highlight some of the biggest threads.

For investors, a return is needed soon, and for many, it won’t arrive for a while yet. For organizations, AI’s promised impact simply hasn’t materialized at scale. And for workers, the fear that AI will eliminate most white-collar jobs appears, so far, overstated.

Things are going to change, but one thing remains clear: AI is going to reshape the future. Regardless of where you sit in the value chain, learning these tools and how to actually implement them inside your organization is a clear way to get ahead.

Like computers and the internet before it, we should treat AI as a tool to move the needle. There are so many ideas out there still waiting to be built. Amid all the fear surrounding AI, try it for yourself. Use it to your advantage. You might be surprised by what you can do with it.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

References

Dubois, N. (2026, March 30). OpenAI business model explained: 900M users, only 5% paying. European Business Magazine. https://europeanbusinessmagazine.com/sam-altmans-openai-is-burning-billions-most-users-pay-nothing-as-anthropic-closes-in/

Field, H. (2025, August 18). MIT report: 95% of generative AI pilots at companies are failing. Fortune. https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/

Fortune. (2026b, June 1). Apollo chief economist says there’s “zero evidence” AI is killing jobs—in fact, he says it’s creating them. https://fortune.com/2026/06/01/apollo-chief-economist-torsten-slok-zero-evidence-ai-killing-jobs-says-its-creating-them/

GAGI: A Gini-adjusted GDP-per-capita index for distribution-aware macroeconomic welfare monitoring. (2026). arXiv. https://arxiv.org/pdf/2606.09944

Dream machine: The next creative economy. (2026). arXiv. https://arxiv.org/pdf/2606.26114

Investing.com. (2026, May 22). The AI token pricing crisis behind OpenAI and Anthropic’s revenue race. https://www.investing.com/analysis/the-ai-token-pricing-crisis-behind-openai-and-anthropics-revenue-race-200680777

OpenAI. (2026, March 31). OpenAI raises $122 billion to accelerate the next phase of AI. https://openai.com/index/accelerating-the-next-phase-ai/

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

However, that simply isn’t the case anymore. AI has transformed from chatbots to autonomous agents, handling core business workflows within highly regulated/confidential workstreams with minimal margin for error.

From this shift, the demand for AI-centered GRC and testing/evaluation platforms has skyrocketed. Just think about it - without the tools to make sure new AI developments follow compliance requirements and assimilate correctly into legacy workflows, new advances in AI would cease to exist.

According to the McKinsey & Company The State of AI Report, enterprise AI adoption has grown from 20% in 2020 to 72% today. As these technologies become more intelligent and useful, and as enterprises find ways to integrate them into their systems, the need for platforms to ensure safe adoption has risen accordingly (2025-26).

That being said, the extent to which AI is currently implemented remains minimal, demonstrating clear bottlenecks to AI’s integration. As companies (and governments) begin to deploy AI agents across different workstreams, these platforms will only become increasingly necessary.

Here are three early stage companies aligning themselves with this tailwind, helping ensure new AI developments align with compliance requirements and are safely deployed.

Subscribe now

1. SolidCore.AI

$4 million, Seed. San Jose, CA. Investors include Runtime Ventures, Epic Ventures, and notable cybersecurity angel investors, with recognition from NVIDIA Inception, Google Cloud Builders, Microsoft Founders Hub, and AWS Activate.

SolidCore.AI is an AI-specific governance and compliance platform that bridges traditional Governance, Risk, and Compliance tools (GRC) with new AI-specific regulations. Traditional GRC tools were not built for model governance, prompt injection risks, and government-specific AI frameworks (i.e., the EU AI Act) - SolidCore addresses this gap, ensuring companies remain aligned with these requirements.

Despite their early stage and limited funding, they have been recognized by all four major AI infrastructure providers: NVIDIA, Google Cloud, Microsoft, AWS, as well as public endorsement from Schellman, the first ANAB-accredited ISO 42001 certification body.

Their founding team is not only experienced but also helped define what the AI regulatory/compliance landscape looks like today. As the United States, EU, and other developed economies continue to implement AI compliance legislation, the demand for SolidCore.AI will continue to rise.

Here's what stands out about SolidCore.AI after running my Second Layer evaluation:

• Founder-Market Fit: Eric Chiu (CEO) previously founded HyTrust, which secured cloud infrastructure for Bank of America, UnitedHealthcare, Honeywell, and U.S. Central Command before being acquired by Entrust in 2021. Hemma Prafullchandra (CTO) served as Microsoft’s AI Security, Privacy, and Data Residency Architect; she defined the company-wide security requirements for all GenAI integrations and Copilots.

• Timing & Competition: Existing GRC tools (Vanta, Drata) weren’t built for model governance, prompt injection risks, or AI-specific regulatory frameworks — creating a compliance vacuum SolidCore is positioned to fill. Mandatory compliance is the strongest possible forcing function for enterprise buying.

• Traction Quality: Recognition from all four major AI infrastructure providers — NVIDIA Inception, Google Cloud Builders, Microsoft Founders Hub, and AWS Activate — is rare at the seed stage and signals SolidCore is positioned as the category-defining player.

2. Freeplay

$8.85 million, Series A. Lafayette, CO. Investors include Renegade Partners, Conviction (Sarah Guo), Matchstick, Next Frontier Capital, PWV (Tom Preston-Werner), and Vermillion Cliffs, with operator-angels from GitHub, Atlassian, Dropbox, Roblox, and Scale AI.

Freeplay provides an LLM testing environment for major AI-driven deployments, allowing product managers, engineers, quality assurance specialists, and domain experts to collaborate on the same evaluation surface. Freeplay allows its customers, which include major corporations like Chime, Raptive, and Maze, to test and improve AI applications against their actual data.

Most LLM evaluation platform tools were built to serve engineering teams; however, Freeplay serves the broader enterprise team. As product teams continue to become increasingly involved in the orchestration and deployment of new AI applications, creating a more product team-friendly application is an extremely sensible play.

Their founding team includes two very successful entrepreneurs, Ian Cairns (CEO) and Eric Ryan (Co-founder), who built the development platform Gnip and sold it to Twitter for $134 million in 2014. Freeplay has already made a major impact on its enterprise customers, who have reported a “10x change in iteration speed” since integrating its technology.

Despite success across my Second Layer metrics, here are a few stand-out factors:

• Founder-Market Fit: Ian Cairns (CEO) and Eric Ryan are repeat successful founders. The team has the rare combination of developer-platform experience and enterprise sales chops that defines the LLM evaluation category’s best operators.

• Investor Signal: $5.6M Series A led by Renegade Partners, with Conviction (Sarah Guo), Matchstick, Next Frontier Capital, PWV (Tom Preston-Werner, co-founder of GitHub), and Vermillion Cliffs participating. The operator-angel base, which includes founders from GitHub, Atlassian, Dropbox, Roblox, and Scale AI, mirrors the customer base Freeplay is selling to.

• Founder Commitment: Four years building toward Series A, with deliberate positioning bet on the cross-functional product team (rather than engineers) as the future buyer of AI infrastructure. Customer testimonials reporting “10x change” in iteration pace after integration confirm the positioning is winning where it matters.

3. Confident AI

$2.2 million, Seed. San Francisco, CA. Investors include Y Combinator (W25), Flex Capital, Oliver Jung, Vermilion Cliffs Ventures, Liquid 2 Ventures, January Capital, and Rebel Fund.

Confident AI is the open-source standard for testing and evaluating LLM applications. Taking the open-source structure that built MongoDB and GitLab, Confident AI now processes over 2 million AI infrastructure evaluations every day and powers LLM deployments for a series of renowned enterprises, including Microsoft, AstraZeneca, and Boston Consulting Group.

Teams can experiment with different prompts, models, and parameters, iterate on AI applications, catch regressions before deployment, and test for safety vulnerabilities within their own AI systems. With such an extensive suite of open-source products, Confident AI boasts over 500 AI companies as customers.

Founders Jeffrey Ip (previously Google, YouTube, and Microsoft AI) and Kritin Vongthongsri have been very conservative in their funding amounts, demonstrating strong capital efficiency despite significant enterprise traction.

Here are the strongest signals within Confident AI:

• Founder-Market Fit: Jeffrey Ip (CEO) brings frontier AI experience from Google/YouTube, and Microsoft AI/Office 365. The combination of working inside the largest AI deployments in the world before building evaluation infrastructure for them is the cleanest founder-market fit in the LLM evaluation category.

• Capital Efficiency: $2.2M raised, 8 employees, 2 million evaluations processed daily, and enterprise customers including Microsoft, AstraZeneca, AXA, and BCG. This is among the strongest capital efficiency profiles across the entire Second Layer pipeline.

• Product-Market Fit: 12.6K GitHub stars, 3M+ monthly downloads, 2M evaluations daily, and four Fortune 500 customers (Microsoft, AstraZeneca, AXA, BCG) at the seed stage. The combination of open-source adoption AND enterprise customers at this stage is very rare, as most seed-stage AI infrastructure companies have one or the other, not both.

Closing Thoughts

The only way new advances in AI can actually be deployed is if they are deployed safely. Traditional sandbox environments were not built for enterprise teams building AI, and legacy GRC platforms were not built for the AI era. These companies are filling a gap that is only likely to widen as AI applications become more advanced (and also, more potentially dangerous if put in the wrong hands).

These companies are helping ensure that the runway for the next wave of AI technology has the track it needs to run smoothly. I would definitely keep an eye on these three over the next few months.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

Reference

McKinsey & Company Global Survey, “The State of AI: Closing the Gap Between Adoption and Impact” / “AI at work but not at scale” (November 2025 / Q1 2026 tracking).

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Website: https://bryan-hanley.lovable.app/

Leave a comment

This milestone is indicative of a much more significant trend: accelerated adoption of clean energy resources. This comes as data centers demand extensive usage of power, with estimates signalling US data center electricity consumption is going to triple from 2021 to 2030, and could consume 9-17% of all US electricity by 2030, up from just 4-5% today (Electric Power Research Institute, 2026; International Energy Agency, 2025). A shift towards cleaner energy solutions will help us meet this demand.

There has been a lot of media attention around this as well. As many lobby against the construction of data centers, energy consumption is often a priority concern for Americans in nearby towns. Undoubtedly, there’s been a significant push for the creation, enablement, and implementation of environmentally-friendly energy solutions to help satisfy a rapidly growing demand.

Nonetheless, what most Americans (and investors) don’t realize is the widening gap between clean energy solutions and their deployment - put another way, the clean energy industry has an infrastructure problem.

Extensive permitting, site selection, and interconnection failures have prevented the mobilization of new technologies into the American energy supply - it’s not a lack of technology, but the solutions to enable its integration. In fact, according to a whitepaper from Paces (featured later!), 70-90% of renewable energy projects stall or fail before construction begins (Paces, 2024).

Turns out, the complications lie in the planning, infrastructure, and rollout - not in the technology, which is why for this week’s Second Layer Spotlight, I’m focusing on companies addressing what I believe truly matters.

Here are three Second Layer-aligned companies helping make clean energy technologies a reality.

Subscribe now

1. Paces

$12.9M, Series A. Brooklyn, NY. Investors include Navitas Capital, Suffolk Technologies, MCJ Collective, and Resolute Ventures.

Paces is an AI-powered software platform that helps clean energy and data center developers locate the best sites for new projects. Paces integrates zoning, interconnection, environmental, and permitting data on individual land parcels. It also provides a Permitting Predictor, using zoning and regulatory data to forecast permitting risk across different jurisdictions.

This helps compress months of manual due diligence into just minutes - it’s significant diligence hurdles that prevent and delay essential projects from taking off in the first place. Paces is enabling faster decision-making at a time when accelerating energy demand calls for it.

Led by James McWalter, employee #1 at a Google-acquired AI startup, as well as Charles Bai, who led AI efficiency and sustainability at Meta, the team boasts engineers from Meta AI, DeepMind, Tesla, TotalEnergies, Con Edison, and GE - a great balance between leading tech and energy companies.

Some areas that stand out about Paces:

• Timing & Competition: IRA-driven renewable acceleration, the data center boom creating new grid demand, and the failure of legacy due diligence tools to scale have created exactly the kind of compounding tailwind that defines a Second Layer opportunity.

• Founder-Market Fit: Two technical founders with frontier AI backgrounds at Google and Meta. The broader team adds direct utility and energy operational experience, which is very impressive for an early-stage startup.

• Product-Market Fit: Customers include EDF Renewables, AES, and Third Pillar Solar, which are three of the most credible developers in US renewable energy. Customers span solar, storage, EV charging, data center, and carbon sequestration developers, demonstrating traction across a variety of hyper-growth customer segments.

2. Reunion Infrastructure

Seed Stage (exact funding undisclosed). San Francisco, CA. Investors include Segue Sustainable Infrastructure and a dozen renewable energy entrepreneurs and CEOs.

Reunion Infrastructure is a marketplace and compliance platform for transferable clean energy tax credits under the Inflation Reduction Act. Reunion became the first company to build a digital marketplace for the trade of clean energy tax credits.

Financing has long been a challenge for renewable energy projects, including solar, battery storage, and geothermal energy. By supporting every step of the transaction process, from identifying opportunities to providing market data and contract negotiation, Reunion helps renewable energy projects get done.

Reunion has overseen the exchange of over $7 billion in tax credit transfers (100+ deals) since 2024, with an average close time of under 45 days from term sheet to signing. Over 95% of signed term sheet deals close on the platform.

With an impressive team led by Andy Moon and Billy Lee, both who have spent careers in solar tax equity and project finance before starting Reunion, they have a ton of experience raising capital to fund global solar projects.

Reunion Infrastructure stands out across a few key areas:

• Founder-Market Fit: Two co-founders who spent their entire careers in solar tax equity and project finance before the IRA’s transferability provisions made tax credits freely tradable for the first time. The team’s collective network across major banks and Fortune 500 tax departments is a significant moat.

• Product-Market Fit: $7B+ in tax credit transactions facilitated since 2024, across 100+ deals. Buyers include Mercantile Bank and Fortune 500 corporations. Q3 2025 was Reunion’s highest-ever transaction volume.

• Capital Efficiency: Profitable. Reached $7B+ in transactions, 100+ deals, and 20 employees on an undisclosed seed round. One of the strongest capital efficiency profiles in clean energy software, a direct parallel to other category-leading bootstrapped energy infrastructure plays.

3. Voltpost

~$10 million, Series A. New York, NY. Investors include RWE Energy Transition Investments, Twynam, Exelon Foundation, and Climate Capital, with a strategic partnership from AT&T.

Voltpost is revolutionizing EV charging stations, retrofitting existing city-owned lamp posts into curbside EV charging stations. Their modular unit attaches to a lamp post’s existing electrical supply without new trenching, new conduits, or new permits for underground work.

The charger can be accessed across public/private lamp posts via an intuitive mobile app, which provides nearby Voltpost chargers, lets you track charging sessions, and lets you manage payments easily.

Cities have millions of lamp posts already connected to the grid, unused for 10-12 hours every night. Voltpost provides a use for these lamp posts while addressing the EV charger supply shortage that has not kept up with EV demand.

Despite their early stage, they have established partnerships with a multitude of notable companies, including AT&T, Eversource, Zipcar, and ComEd.

Despite performing well on all criteria, here’s what particularly stands out about Voltpost:

• Market Size: Approximately 115 million lamp posts in the US, and every urban municipality with EV adoption mandates is a potential customer. EV mandates in California, New York, and across Europe create structural demand drivers that compound annually. 40-50% of US urban households have no driveway and no home charging option.

• Timing & Competition: Federal NEVI funding ($7.5B) is largely directed at highway corridors, creating a curbside funding gap that Voltpost fills. DC fast chargers cost $50K–$150K per unit and require grid upgrades; Voltpost units deploy in hours using existing electrical supply at a fraction of the cost.

• Investor Signal: RWE Energy Transition Investments (the venture arm of one of Europe’s largest renewable energy utilities) led the seed round, signaling sector-specific validation rare for a hardware-first startup. The AT&T partnership (November 2024) confirms operational scale through a Fortune 500 connectivity provider.

Closing Thoughts

Clean energy is the future of energy - but the only way to bring clean energy to life is through more effective planning, better financial infrastructure to fund them, and solutions that integrate effectively into existing systems/grids. Solely focusing on the technology is shortsighted - building the infrastructure to enable new technologies and projects from taking place is where investors should be looking.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

References

Electric Power Research Institute. (2026). Powering intelligence 2026: Analyzing data center load growth and grid impacts. EPRI. https://powering-intelligence.epri.com/load-growth.html

Ember. (2026, June 10). Solar overtakes coal in US electricity for the first month on record. https://ember-energy.org/latest-updates/solar-overtakes-coal-in-us-electricity-for-the-first-month-on-record/

International Energy Agency. (2025). Energy and AI. IEA. https://www.iea.org/reports/energy-and-ai

Paces. (2024, July 24). Paces raises $11 million in Series A funding to accelerate renewable energy projects. Paces. https://www.paces.com/news/paces-raises-11-million-to-accelerate-clean-energy-development

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

The United States spends $14,885 per capita on health, more than double the average for OECD (Organization for Economic Cooperation and Development) countries of $5,967 per capita (according to the OECD Health at a Glance Report). That’s $5.5 trillion a year on healthcare alone.

Logic would suggest the U.S. would be the healthiest country - this assumption, however, is far from the truth. The Commonwealth Fund did a report on 10 of the wealthiest nations, ranking healthcare effectiveness across 5 different key metrics: access to care, care process, administrative efficiency, equity, and health outcomes. The U.S. scored last on 3 out of 5 of these metrics (including health outcomes), earning the lowest spot in the aggregate rankings.

The issue is clearly not a funding problem - it’s an operational problem. And it needs to change.

We all know someone (or many) who currently or previously needed significant healthcare treatment - when it comes to friends and family, you do the very best you can to give them the best possible care. But whether it’s prolonged time-to-treatment, insurance billing complications, or outrageous medical bills, millions struggle to get the care they deserve.

Powered by AI, many startups are trying to change the narrative. America has the best schools and best hospitals in the world, and our healthcare system should reflect this. America doesn’t need more doctors: it needs better systems around those doctors, helping them operate as they were intended to. This is the Second-Layer focus.

Here are three Second-Layer aligned companies seeking to change the status quo and make healthcare more efficient for millions of Americans across the country.

Subscribe now

1. HEARTio

$5.28M, Seed Stage. Pittsburgh, PA. Investors include Shepherd Ventures, Intelligence Ventures, Audacious Capital, VU Venture Partners, LifeX, and Bessel.

Coronary artery disease (CAD) is the leading cause of death in the US, killing an estimated 375,000 Americans every single year. Diagnosing the disease remains a critical issue: coronary CT angiography and catheterization labs are expensive and geographically concentrated in certain urban areas.

Additionally, when primary care physicians highlight heart conditions, it’s difficult to distinguish between milder heart issues and more critical ones (like CAD). This leads to misdiagnoses, where patients with active disease aren’t treated fast enough, and patients without disease are unnecessarily over-referred for costly medical treatments.

HEARTio provides AI-powered software layered on top of existing ECG infrastructure, which more accurately detects coronary artery disease on a per-vessel basis. This enables more accurate disease detection (clinical accuracy over 90%), pinpointing which specific coronary arteries are blocked (in just 30 seconds) and helping distinguish between CAD and other milder issues.

With FDA Breakthrough Device Designation, two peer-reviewed publications in the Canadian Journal of Cardiology, and JACC Advances with 16,000+ patients, HEARTio is a cost-effective, proven solution to one of America’s biggest problems.

HEARTio stands out across the following areas:

• Technical Execution: FDA Breakthrough Device Designation in 2020, 16,000+ patient cross-validation, and a 510(k) premarket study in the works show that they are on their way to commercialization. 3 issued patents as defensibility.

• Founder-Market Fit: 3 University of Pittsburgh PhDs (led by Utkars Jain, CEO) who built the technology during their doctoral research, with an experienced advisory board to support go-to-market.

• Market Size: there are 14 million ECGs in the U.S. alone - this software can be layered onto these existing ECGs, demonstrating simple integration.

2. Kinstead Health

Pre-seed stage (funding undisclosed). New York, NY. Backed by Wisdom Ventures, Headstream Innovation, and angel investors.

There has been a consolidation of our healthcare system: the share of physicians in independent practices fell from 60% in 2012 to under 25% today. Nurse practitioners are the fastest-growing clinical workforce in the country; however, as founder Taylor Rose realized while working at the Children’s Hospital of Philadelphia, burnout in major health systems is very common, often due to a misalignment between preferred and mandatory working hours.

Kinstead Health is an AI-native platform that helps nurse practitioners (NPs) launch and grow independent practices. They handle setup, scheduling, documentation, billing, credentialing, compliance, and growth all under a single umbrella. With Kinstead, NPs go from idea to practice in 90 days or less.

NPs who want to go independent face an array of challenges that Kinstead supports them with, inspiring the quickly growing number of NPs to take the leap and start their own independent practice.

Here are a few areas where Kinstead excels:

• 1A Founder-Market Fit: Taylor Rose (McKinsey, CHOP, Roivant Social Ventures, McKinsey/WEF women’s health report) and Parth Chodavadia, MD (ex-PathAI Digital Diagnostics); the team has lived the problem from hospital administration, consulting, and clinical practice angles simultaneously

• 3B Timing & Competition: With NPs growing so quickly, the need for an AI-native platform is immense, especially as burnout rates of NPs in large systems remain sky high. Other teams lack the direct operational and commercial experience of Kinstead.

• 3A Market Size: Scope-of-practice expansion in 27+ states; every independent NP practice in the U.S. is an addressable customer, and the trend is accelerating (NPs are the fastest growing health practitioner occupation).

3. Understood Care

$8.4M, Seed Stage. New York, NY. Investors include Y Combinator, 1984 Ventures, Rethink Education, and Zeal.

Medicare Advantage covers over 50% of all Medicare beneficiaries, which equates to over 31 million Americans. While these plans were designed to handle care more efficiently than traditional Medicare, there have been a multitude of issues associated with them: an explosion of authorization requests, claim denials, and administrative friction.

The largest payers (UnitedHealthcare, Humana, Aetna) deny roughly one in five prior authorization requests, and most beneficiaries simply give up rather than appeal their cases. This means many Americans don’t receive the care their plans should theoretically cover.

Understood Care provides AI-native Medicare patient advocacy, helping Medicare beneficiaries navigate Medicare Advantage plans, file appeals, manage authorizations, and coordinate care across multiple providers.

2024 Centers for Medicare & Medicaid Services (CMS) Codes are brand new, allowing patient advocacy to be billed directly to Medicare for the first time. Understood Care was built specifically around these codes, where CMS pays for the service directly. By becoming the first venture-backed entrant to structure their revenue in this way, they are establishing a strong, defensible moat that must be emphasized.

Here’s what specifically stands out in my assessment of Understood Care:

• 2B Revenue Signals: CMS-funded revenue from day one via 2024 advocacy billing codes; Medicare itself pays for the service, eliminating customer acquisition burn typical of healthcare startups

• 1A Founder-Market Fit: Sam Wu, former senior operator at Cedar (healthcare billing) and co-founder of Finni Health (YC W23, pediatric autism care); repeat founder with directly applicable domain experience is an unusually strong signal at the Seed Stage.

• 3B Timing & Competition: Medicare Advantage now covers 50%+ of all Medicare beneficiaries with prior authorization and denial rates at all-time highs; 2024 CMS advocacy codes are new, and Understood Care is the first major venture-backed company built around them.

Closing Thoughts

U.S. healthcare is a problem, but I believe companies like the ones highlighted above are truly making a difference. AI gets a lot of hate (and sometimes for the right reasons). However, seeing the impact AI has had as a transformative tool, empowering the above founders to bring truly great products to healthcare, gets me genuinely excited. I can’t wait to see the impact these founders have on the world.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

This week, I want to focus on fintech: more specifically, those building the future of finance, as well as those building systems, processes, and technology to protect it. These companies provide the infrastructure, governance, and safety net essential for other services to properly function.

As global finance has become more integrated, it has also become more complex: these SecondLayer-aligned companies are helping ensure financial platforms have the proper infrastructure to operate effectively, while also helping keep the bad actors out.

Here are three early-stage companies within fintech, payments, and financial compliance that are helping drive a more integrated, safer financial future.

Subscribe now

1. Flagright

$7.6M, Seed Stage. Decentralized HQ (San Jose / London / global hubs). Investors include Frontline Ventures, Moonfire Ventures, Y Combinator, Pioneer Fund, and angels.

As global finance becomes more integrated across a variety of different platforms, including neobanks, payment processors, and crypto firms, the demand for financial compliance has grown substantially.

New financial technology applications need anti-money laundering (AML) platforms on day one, especially as AI exacerbates the threat of cybercrime and fraud. Flagright alleviates this threat, building the layer that makes mandatory AML infrastructure functional.

Flagright provides real-time transaction monitoring, AML screening, and case management through an API-first, AI-native platform that integrates in just weeks, not months. The speed of integration provides a significant moat in a market that continues to grow significantly.

Flagright stands out the most within the following criteria:

• Founder-Market Fit: Baran Özkan, CEO, and Madhu Nadig, CTO, are experienced operators within the financial compliance space.

• Timing: a series of crypto and fintech legislation, including the GENIUS Act and CLARITY Act, demonstrates Flagright is scaling at the right time.

• Traction Quality: 50+ customers across 30+ countries with $4M+ ARR with just 27 employees is phenomenal.

2. Parcha (building GREP AI)

$5M, Seed Stage. Based out of San Francisco. Investors include Kindred Ventures, Initialized Capital, Propel, Fin Capital, Liquid2 Ventures, and angels.

Parcha perfectly fits the Second Layer thesis: as fintech continues to scale globally, so does the need for compliance standards within these companies. Parcha builds AI agents that sit inside compliance workflows, automating document review, risk management, and decision-making that empowers compliance teams.

Unlike competitors (most notably WorkFusion), Parcha provides operational AI agents that support speed and accessibility for compliance teams. Their UX is simple and modern, making it easy for compliance teams to integrate and manage these agents within existing systems.

When evaluating Parcha as an early-stage investment opportunity, a few areas stand out:

• Founder-Market Fit: Cofounded by AJ Asver (CEO) and Miguel Rios-Berrios (CTO), both formerly of Brex, lived the financial compliance problem directly.

• Market: Fintech revenue is projected to reach $1.5 trillion by 2030: the firms driving this revenue will need the compliance infrastructure to support their continued growth.

• Investor Signal: Kindred Ventures and Initialized Capital both have very strong fintech infrastructure track records. Kindred (small but impressive portfolio, leading the pre-seed and seed stage investments for Parcha), along with Initialized Capital, which invested in Coinbase and Reddit early.

3. Themis

$9M, Seed. Based out of New York. Investors include TTV Capital and The Fintech Fund, with additional backing from angel investors across the banking and fintech compliance community.

Themis focuses on a different Second Layer dynamic when compared with the other two: the rise of banking as a service and the need for network management as banks begin to sponsor numerous fintech platforms.

Managing the web of relationships between banks and fintech platforms with pre-existing solutions is simply unsustainable today. While competitors focus on bank-fintech compliance relationships from one side, Themis has built a shared workspace for both sides to operate simultaneously.

Themis excels across a variety of different early-stage criteria, but most particularly in the following areas:

• Founder-Market Fit: Neepa Patel is an experienced operator within the financial compliance space. She worked as a bank regulator at the OCC (Office of the Comptroller of the Currency) before transitioning to compliance roles at Morgan Stanley, Deutsche Bank, and r3.

• Early PMF: 10x revenue growth within Governance, Risk Management, and Compliance (GRC) is quite impressive, as GRC is traditionally very difficult to sell for early-stage companies.

• Timing: Between 2022 and 2024, the OCC and other federal regulatory bodies issued a wave of consent orders, enforcement actions, and guidance letters targeted towards sponsor banks of fintech companies. These were issued due to a lack of adequate oversight over these fintech partners, validating the need for the solution Themis provides.

Closing Thoughts

The need for companies to ensure the safe deployment and functionality of AI has continued to grow, particularly as models and agents become increasingly advanced and complex.

The importance of Founder-Market Fit is clear, particularly within fintech/financial compliance. Having experienced the complications first-hand is essential to building a product that the market truly needs.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

What Really Matters

Thank you for tuning in this Memorial Day - a day that belongs to the brave Americans who give people like me the freedom to write and to learn. We owe it all to them.

As always, if you have any questions, feedback, or recommendations, I’d love to hear them. Please don’t hesitate to reach out to me via email or LinkedIn.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

I was reading one of my favorite collections of venture capital theories, perspectives, and lessons, which are derived from concepts within economics, math, physics, chemistry, biology, and psychology. This document was authored by Chris Paik, General Partner and Co-founder of Pace Capital (and former partner at Thrive Capital, early backer of Twitch, Patreon, and Unity).

If you (like me) never thought of the role other industries/areas of study could play in how startups and the businesses/consumers they serve operate, this collection does a fantastic job of explaining the connection. Understanding why people and businesses operate the way they do is perhaps more important than the economics itself. The collection does a great job of explaining how user behavior creates the demand fulfilled by entrepreneurs.

Subscribe now

Being the Answer to ‘Why Now’

One thought that resonated with me most was a startup being the answer to the ‘why now’ question for others companies. In other words, providing the runway for other companies to launch from. If your idea can deliver an answer to this ‘why now’ question for others, then it’s capable of becoming the unicorn VCs dream about. This is especially true during a major technological breakthrough, where there is a ton of untapped value potential within a rapidly-growing market.

A perfect example of this is Shopify: a startup making it easy for individuals and businesses to build, run, and scale their own business. It took advantage of the rapidly growing e-commerce market and the rise of early social networks, serving as the platform of choice for numerous well-known startups including Gymshark and Fashion Nova.

Importance of the Market

For those unfamiliar with investment pitches, the following three terms are used to reflect the attractiveness of the market: the Total Addressable Market (or TAM), which is simply the value of 100% of the revenue within the market; the Serviceable Addressable Market (or SAM), which is a component of the TAM you realistically can capture; and the Serviceable Obtainable Market, which reflects what a startup can realistically capture within the short term (1-3 years). If a startup is building in an industry where these market sizes aren’t favorable, it certainly hurts its investment attractiveness.

As Chris highlights in the document, venture capital is built to serve a single purpose: deliver capital to its limited partners by investing in areas capable of explosive value creation within a certain period of time. When a major market shift occurs (i.e., the AI-Era), the opportunity for new ventures to scale and capture the value within the industry is seismic.

However, as more and more founders innovate, the amount of capturable value declines over time (similar to the decay of a radioactive isotope). This is why it’s so important for startup founders to have an answer to the “why” in the “why now”: what’s so special about this time in history, what’s so special about the market, and why am I (the founder) poised to capture a lot of its value?

Paik provides a good example of this: today, we rarely see the creation of major mobile-first social companies in 2026. The smartphone market matured (at least in the Western World) around 2014 - 2016. Snapchat, Instagram, TikTok, and others captured a majority of this market; today, there simply isn’t enough “uncaptured value” within the mobile-first social market for a new startup to be an attractive investment. For a venture capitalist, the market simply won’t cut it.

Why Now is Such an Exciting Time

Today, it’s a bit different. AI in its current form is in its breakthrough stage, hence the significant amount of VC investment that’s been committed to it; a study from the Organisation for Economic Co-operation and Development found that artificial intelligence firms accounted for 61% of global venture capital investment in 2025 (OECD, 2026). However, the next question has to be, what’s next?

Yes, the shift from AI as a tool to the implementation of AI into workflows (agentic workflows, to be specific) is another example of explosive value creation. And while I do think we have some time before the AI-first era reaches its maturity from a value creation standpoint, at some point a lot of that value will be captured. I believe taking the Second Layer approach to venture investing could give us insight into what the next areas of explosive value creation could be.

Like Chris, I look at venture investing as an open field of value waiting to be captured by startups and investors alike. But what happens when all the value is captured for AI in its current form? Where do we look next? By focusing on the second layer, investors face less competition from other VCs across areas with a ton more untapped value to capture.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

References

Organisation for Economic Co-operation and Development. (2026, February). AI firms capture 61% of global venture capital in 2025. OECD. https://www.oecd.org/en/about/news/announcements/2026/02/ai-firms-capture-61-percent-of-global-venture-capital-in-2025.html

Paik, C. (2021, January). Frameworks v0.2. Pace Capital. Retrieved May 19, 2026, from https://docs.google.com/document/d/1-UiEeoiV0xBFVZgid63FRaph03OCmHzyEExubn63j0U/

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

This week, the second-layer vertical of focus is AI infrastructure. A lot of investor and media attention remains on developing a more powerful and capable LLM, or finding new ways to implement AI into existing workflows for operational improvement.

However, AI systems are only as effective as the infrastructure in place to support them: that means enabling integration into systems, understanding and addressing bottlenecks, and digesting unstructured data across a variety of sources.

It won’t be an absence of innovation that will inhibit the development of new AI-focused technologies - it will be limitations in the infrastructure meant to support this growth in innovation at scale.

Here are three early-stage companies within the AI infrastructure vertical with promising early signals:

Subscribe now

1. Arcade.dev

$12M, Seed Stage. Based out of San Francisco. Investors include Flybridge, Hanabi Capital (Mike Volpi), Neotribe Ventures, Garuda Ventures, and angel Andy Rachleff

Founded in 2024 by Alex Salazar and Sam Partee, Arcade AI is a tool-calling and authentication infrastructure platform for AI agents. AI agents need to call external tools, whether it’s sending emails, querying databases, reading calendars, or hitting APIs.

However, each one of these activities requires secure authentication, permission scoping, and token management - AI agents constantly run into bottlenecks across these areas, inhibiting workflows from running smoothly. Arcade AI builds the layer that handles this, giving agents pre-built connections to dozens of services alongside a Software Development Kit.

Arcade.AI scored the highest across Founder-Market Fit (previous founder exit within the space), TAM (agentic software market is massive and growing), Timing (enterprise agentic AI adoption is scaling exponentially), and Investor Signal (funding partners have an extensive, successful track record).

2. Trellis AI

$500k Y-Combinator Raise. Based out of San Francisco. Funded by General Catalyst (undisclosed amount)

Trellis AI automates healthcare paperwork and prior authorization workflows for healthcare providers. It builds AI agents that process unstructured clinical documents, including referrals, medical charts, insurance guidelines, and appeals, converting them into structured data directly integrated into Electronic Health Record systems.

By enhancing these processes, the time-to-treatment bottleneck is significantly reduced, allowing patients to be treated more quickly while allowing health centers to treat a higher volume of patients.

The U.S. spends over a trillion dollars a year on administrative healthcare costs, delaying care, draining provider revenue, and driving staff burnout. By reducing time-to-treatment by over 90%, Trellis is tackling one of the most substantial problems facing American families today.

Trellis AI performed exceptionally well across a few areas: backing from General Catalyst demonstrates high investor quality, Product-Market Fit (billions in annual therapy processing, nationwide scale across hundreds of locations), and Technical Execution (clinical training on millions of data points, along with integration into EHR platforms).

Though Trellis AI is early stage, they have already experienced quite significant, noteworthy traction. It is a company I’d keep an eye on.

3. Maxim AI

$3M, Seed Stage. Based out of San Francisco. Seed round led by Elevation Capital, with Investment from various Angels: Founders of Postman (Abhinav Asthana), Chargebee (Krish Subramanian), Groww, Razorpay, and Media.net

Maxim AI is an end-to-end evaluation and observability platform for enterprise AI agents and LLM applications. Most enterprises managing LLM applications use an observability tool or an evaluation framework separately - Maxim was developed as a unified solution, targeting teams that have adopted AI implementation at scale and need their systems evaluated for quality assurance.

Maxim AI covers Pre-deployment (testing LLMs before production), Post-deployment (live monitoring, tracing, and observation), and Quality assurance (evaluation of outputs to detect failures throughout the complete lifecycle). By bridging the gap between observability and evaluation, Maxim AI eliminates the need to toggle between multiple tools.

Where Maxim performs exceptionally well in my startup analysis is in the Market (every enterprise AI deployment needs quality assurance infrastructure), Timing (LLM experimentation to production is happening fast), and Investor quality (Elevation Capital is a well-established fund).

Closing Thoughts

I’m always happy to share more specific details about these startups and why they are on my radar. I plan to share high-quality early-stage startups across a new vertical each week - if there’s a vertical that excites you, please let me know.

As always, if you have any questions, feedback, or recommendations, I’d love to hear them. Please don’t hesitate to reach out to me via email or LinkedIn.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

Subscribe now

The Issue with Venture

VC investing, especially compared to more easily accessible means of investing such as index funds, brings its fair share of problems. The first of note is that all VC firms aren’t created equal; the extent to which these firms differentiate amongst each other from a return perspective is quite significant. Many top-performers (think Sequoia, Andreesen-Horowitz, General Catalyst) perform in the range of 15% – 27% IRR annually, providing LPs with significant returns on their investment.

However, not every firm is like Sequoia (though they strive to be). In reality, it’s these top firms that are going to have their top pick of the very best founders: they have the track record, they have the capital (happy LPs will continue to invest into new fundraises), and they have the expertise. But what about the firms that don’t perform as well? Smaller firms without a track record, or new funds/inexperienced GPs? The story is a bit different.

The data is clear: top funds perform exceptionally well in terms of IRR, with bottom (and even median) funds performing worse than the S&P 500. Put simply, VC is hard. Between the likelihood of investing in a unicorn to make up for those who don’t make it (especially if you’re a smaller fund), and the associated management fees, expenses, and carry these firms payout back to fund managers, VC isn’t the most lucrative investment option for most. And that’s not even considering the illiquidity of VC investments, where realizing gains from VC funds could take as long as 7-10 years (or longer).

Why People Still Invest in Venture Funds

So if you’re accredited, why invest in venture? Well, it does come with its perks. If you do choose correctly, you will see significant upside, particularly if you agree to favorable terms (liquidity preferences, preferred stock, etc.) By investing in a firm, you also have the opportunity to help shape its course: venture firms sit on the board of directors of the company, and have a major say in the overall strategy of the company you back. You can directly influence their likelihood of success, a path that benefits all parties involved.

With venture investing, you can not only realize significant returns, but you can also help change the world for the better. I certainly recommend giving Ron Levin’s Higher Purpose Venture Capital a read when you get the chance. The book does a really great job of demonstrating how venture-backed founders are effectively addressing the world’s most critical issues in exciting new ways, from poverty to income inequality and education. Most founders really are out to make a difference in the world with their idea, and by investing in them, venture firms are empowering startup founders to do so.

Lastly, venture capital is exciting. It may take years for you to realize the returns of your investments, but the opportunity to make an impact on a company at its earliest years is part of the package, and can’t be dismissed. It’s part of the reason why venture is so competitive career-wise — the opportunity to find the next wave of world-changing companies, and grow alongside them, is very rewarding.

So does venture make sense? Well, there’s a reason why you must be accredited in order to invest in these companies (though ETFs like XOVR and closed investment funds like USVC are trying to change that): there is significant risk associated with these investments.

With 75% of startups failing to return invested capital, venture capital can’t be the money you use to pay the mortgage. But for those willing to take on the risk, and especially those with a successful approach to choosing winners, it could be a high-yielding investment vehicle, providing investors the opportunity to shape the world for the better.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

For decades, space was reserved for government-affiliated organizations, acting as a sort of “geopolitical theatre” to demonstrate navigational capabilities and power. Commerce wasn’t the goal; dominance was, and only the strongest countries in the world had the capabilities of space travel. From Sputnik in 1957 to 2010, fewer than 50 companies globally had any meaningful commercial space activity. Since 2010, that has certainly changed quite a lot.

This change primarily came from SpaceX, which had its first successful commercial launch in September 2008 after the fourth attempt of its Falcon 1 rocket. However, the most monumental shift came from its Falcon 9 Launch in 2017, representing the first successful reuse of a rocket.

This successful reuse not only meant a major shift in how we travel to space but also demonstrated that space travel could be financially viable. What was traditionally reserved for major governments was now accessible for startups - even a startup of $5 million could put a satellite into orbit.

Space tech companies raised $8.17 billion in 2025, a 154% increase from 2024. Startups are realizing the economic potential of space, and investors are taking advantage, particularly within defense and AI. SpaceX enabled a data layer that opened a window of opportunity for space innovation, creating a major second-layer opportunity. And it’s these opportunities that I’m most excited about from an investment perspective.

Subscribe now

Space Startups Aligned with the Second Layer Thesis

As I’ve continued to deploy my Second Layer framework across a variety of sectors, space remains one of the most lucrative and exciting industries, with some incredible founders building within it. Here are a few startups building within the space sector in alignment with the Second Layer Approach that have all demonstrated significant growth and potential, as evidenced by tremendous progress and impressive valuations:

Aalyria

Aalyria, which recently raised $100 million to reach a valuation of $1.3 billion, provides a network orchestration platform that enables stronger wireless internet connectivity for government defense agencies, satellite operators, and commercial aerospace companies by mitigating network errors caused by weather disruptions, atmospheric interference, and shifting military priorities. There are over 11,000 satellites in orbit - when data travels across a satellite constellation, it must move across these various satellites, which are moving at 17,000+ mph, all in different directions. Spacetime reconfigures an internet constellation’s antenna to the path of least resistance, helping ensure critical data is best routed for priority missions.

Privateer Space

Privateer Space led a Series A of $56.5 million, led by Aero X Ventures, in May of 2024. Founded by Apple Co-founder Steve Wozniak, CEO Alex Fielding, and Dr. Moriba Jah, Privateer Space provides real-time debris and collision-risk visualization to commercial and government clients through their Wayfinder platform (which operates as the “Google Maps” of space). Wayfinder enables space object tracking, anomaly detection, space traffic management, and more. As mega-constellation proliferation (the rapid development of hundreds to thousands of small, low-cost satellites in Low Earth Orbit), created orbital congestion, Privateer Space is providing the technology to help operators navigate the crowded space these constellations occupy.

Orbit Fab

Orbit Fab closed a $28.5 million Series A round in April of 2023, led by 8090 Industries, alongside $21 million in defense contracts. This brings their total venture funding to over $45 million. Orbit Fab’s mission is to build the in-space propellant supply chain through their refueling interface, RAFTI, which enables on-orbit refueling with a built-in cooperative docking capability. RAFTI eliminates the need for single-use spacecraft - its refuel design enables spacecraft to get to orbit faster and do more while they’re there, minimizing the supplemental weight associated with traditional fill and drain components. As satellite constellation growth made replacing satellites more expensive than extending their lives, Orbit Fab provided the technology to make this happen, demonstrating direct alignment with the Second Layer thesis.

Unseenlabs

Unseenlabs was founded by three brothers, Clément, Jonathan, and Benjamin Galic, and focuses on space-based radio frequency detection for Maritime Domain Awareness (the effective understanding of anything associated with the global maritime domain, which includes vessels, cargo, people, or infrastructure, and how these could impact security, safety, and the environment). Unseenlabs has raised €120 million, which includes a €85 million Series C round in February 2024. Their Second Layer alignment is clear: satellite proliferation created the infrastructure that enabled the detection of Radio Frequency signals from orbit, which exposed a gap in massive maritime surveillance. Unseenlabs built a constellation of shoe-box-sized satellites that detect ships even when they turn off their AIS (Automatic Identification System) transponder. This provides an additional layer of safety and security, enabling governments and organizations to geolocate any vessel at sea regardless of whether it wants to be found, in any weather, day or night.

Note: The views expressed in this article are my own and are intended for informational purposes only. Nothing here constitutes financial or investment advice. Venture capital involves significant risk, and all investment decisions should be made in consultation with a qualified financial professional.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

A Correlation Ventures study examining over 27,000 investments from 2009–2018 found that 64% of VC deals did not even return the original principal (Berman 2018). Investors know this when they raise their fund and choose investment opportunities. However, it’s the upside on the few investments that provide the substantial returns realized by top venture capital firms (it’s important to note the stark difference in returns of the top-quartile and lower-quartile VC firms from a returns standpoint, which I’ll discuss in greater detail in a different article).

Subscribe now

Two Distinct Approaches to Venture Investing

I had the opportunity to meet with two different VC firms recently, both of which execute different investment strategies influencing both risk and potential returns along different time horizons. The first focused on a high volume of investments, typically as a co-investor (meaning investment rounds aren’t led and investors commit less capital for a smaller stake); the second, a concentrated, hands-on approach to investing (where deals are led and a greater equity stake is obtained).

With the high volume approach, deals are made much more frequently: however, on a per-deal basis, the committed capital is lower because lower equity stakes are secured. This means much less (or even no) board representation, demonstrating less involvement in key decisions of the business, and a reliance that the founders and other investors with a greater stake will make decisions in alignment with yours. On the other hand, the concentrated investment strategy commits capital more carefully, taking a larger equity stake in companies but keeping a much smaller portfolio.

The economics of these two different approaches are quite interesting: when considering a 5-year, 10-year, 15-year, and 20-year horizon, each strategy yields different Net IRR%, different multiples on invested capital, and different levels of risk and liquidity. The below graphs depict different scenarios of capital invested across the two different strategies, and do not represent historical evidence (though, they are derived from analyses of historical performance).

Net IRR Performance

As demonstrated in Figure 1, the concentrated hands-on approach has a greater volatility, commonly referred to as a Beta (ꞵ)*. This is due to the lower volume of investments, and is very similar to why stock indexes are so popular: the greater the number of investments (or stocks), even if a few fail, your returns aren’t influenced as considerably. However, the same goes for your returns: even if a couple of investments perform very well, due to your lower concentration and lower equity stake in these winners, that high return is not as visible.

*For the methodology and source of the data in the below graphs, please refer to the block quote at the end of the article

Figure 1: Annualized Net IRR Over Fund Life

Multiple on Invested Capital

The story for MOIC is a bit different: the larger equity stake leads to a higher multiple for invested capital for the concentrated approach, and this is true for both the median and the best-case scenarios. The best way to describe this is in the case of a winner: if a fund owns a 25% equity stake in a company that reaches a $500 million valuation, that position returns $125 million. For a co-investor in the same deal, if they own just 3% of that same company, they only get $15 million out of it.

Other factors include risk (higher risk tolerance = greater gains), but liquidation preferences and anti-dilution play a role as well. If you are leading deals, you set the path forward, which becomes increasingly important as liquidation preferences and cap tables become more complex. With anti-dilution, as lead investors set the standard, they almost always negotiate pro-rata rights to maintain their larger equity stake, which smaller co-investors may not have.

Figure 2: Multiple on Invested Capital

Risk and Liquidity

Lastly, it’s essential to discuss risk and liquidity. A concentrated profile has a higher risk of loss; as we remember the stats provided earlier, 64% of deals do not even return the principal. It’s equally important to note where the gains actually come from — Dave McClure, founder of 500 Startups, noted that just 1–2% of investments yield the 50x returns investors hope for when they agree to fund a deal (McClure 2015). Finding a unicorn is not only difficult, but when you have a limited number of investment deals to make, the search becomes even more difficult.

Figure 3: Risk and Liquidity Profile Comparison

Conclusion: Both approaches have their own benefits and drawbacks; investing in many companies, just like stocks, reduces your risk; however, given lower equity percentages, you don’t see as much of an upside on your winners. If you’re building a concentrated portfolio with a greater equity stake, influence via board seats, and greater time devoted to helping your portfolio companies grow, you’ll ultimately see more significant returns if these companies scale as hoped. But as we noted above, finding a unicorn (as the name suggests) is no easy feat. That being said, if executed successfully, both strategies have the potential to return substantial gains.

While investment frequency is an important component of fund strategy, there are so many different factors which determine risk and reward that we have yet to discuss. Some of these factors include lead investor success, specific investment areas and stage, and investment terms. I look forward to diving into these areas with you soon.

Note: Illustrative composites synthesized from Cambridge Associates US VC Index benchmarks (2,699 funds, 1981–2025), Carta fund performance data (2,835 funds, 2017–2025 vintages), AngelList LP portfolio analysis (10,665 portfolios), CEPRES database (1,840 funds, 28,452 deals, 1980–2023), Correlation Ventures (27,000 investments, 2009–2018), NBER research on venture capital risk and return, and CAIS/Preqin persistence studies. These are not predictions or backtests of any specific fund. These are median representations of the data, not necessarily demonstrating outliers. All returns net of standard 2/20 fee structures. Past performance is not indicative of future results.

References

Berman, Seth. 2018. “Venture Capital: No, We’re Not Normal.” Medium, May 30, 2018.

McClure, Dave. 2015. “99 VC Problems But A Batch Ain’t One: Why Portfolio Size Matters For Returns.” 500 Hats, May 22, 2015.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

The enhancement of AI applications over the past few weeks has been nothing short of remarkable (and, frankly, a bit scary). The time it takes to move from idea into action continues to decrease substantially, with platforms generating accurate code (which previously would take weeks to months) in a matter of seconds. As many are concerned about the impact these platforms may have on the job market, one clear benefit is this: there’s no more barrier to building that app you dreamed of.

You don’t need to invest hundreds of hours into learning how to code, or scanning websites and research materials to help you define the competitive landscape, or building a PowerPoint presentation from scratch. AI eliminates those barriers, centering the focus on scalability as opposed to early-stage planning and development.

However, with the barriers to entry removed for developing these applications (at least in their early stages), how do you differentiate? It’s no longer about whether or not you can build it, but whether or not you can validate it, scale it, and defend it.

Yes, it’s easy to come up with an MVP for an idea you’ve been thinking about. But the efforts needed to refine, test, and scale aren’t remediated by AI efficiencies. In other words, you still need to build something customers actually want (not what an LLM might indicate they want).

You still need to know how to improve upon your ideas based on feedback from current/potential customers. And you still need to create a “moat” that gets both customers interested in paying and investors interested in writing you a check (if you choose to go that route).

Subscribe now

So how do you create and sustain a competitive moat in the AI-era? As a startup operator, here are some of my thoughts:

Network Effects

Think about Facebook: their primary moat throughout their startup phase was their user base. Demonstrating a growing user base is one of the best forms of defense you can have as you scale your product. People only have so much time in the day, meaning they can only be on so many different applications. That barrier has not changed. So creating a product that becomes more valuable as more users join*, and encouraging users to love it so much that they share it with their friends and family, are keys to scale that provide differentiation and an eventual, sustainable moat.

*I suggest reading The Cold Start Problem: How to Start and Scale Network Effects by Andrew Chen - it does a fantastic job of explaining this concept practically.

Customer Feedback

Speak to your users… but more importantly, listen to your users. I was listening to the Y-Combinator Podcast on how to get your first customers in 2026, and the emphasis of the video was to launch early with a Minimum Evolvable Product (MEP). In launching early, you get invaluable feedback from your first users that can help you shape the course of product development. The human touch must not be undermined: they are the ones who will be buying it. Avoid analysis paralysis, and talk to people who just might like what you’re building.

Barrier by Association

Probably the most straightforward way (though not easy by any means) to protect your product is by association: what’s a company, government, or organization others trust? How could working with them, or obtaining a certain seal of approval or trust mark from them, lead customers to trust you, too?

One way to establish this moat is through defensibility by association (for example, a fintech application launches on the American Express network). The second, by obtaining the legal requirements to operate in areas others can’t. For example, at the startup I work at now, one of our main competitive moats for our conformance testing service is our certification from the global standards provider of traceability. Not everyone can nor will obtain either of these moats, so make obtaining them a priority.

Effective Storytelling

At the end of the day, AI can build a lot - however, it doesn’t tell the story. That part is up to you. Knowing your product, competitors, and the market inside and out (enough to defend it when an investor or customer questions you) is essential. This also highlights the importance of a deep interest in your idea - as Steve Jobs once said, the only way to do great work is to love what you do. Having immense passion for an idea or an industry is essential - not everyone will, and it’s essential for effective storytelling*.

*Many venture investors, particularly early-stage ones, consider passion a very integral factor when considering an investment. This is especially the case if the founder is young or inexperienced —how can a genuine interest or connection with the problem propel them to solve it in a unique or proprietary manner?

Continuous Improvement

Finally, implementing continuous improvement across all the different aspects of your business is essential. Want to know if you should add feature one or feature two? Or both? Ask your users. Want to know if you should extend your offerings to a larger TAM, or focus on a highly specialized niche? Map out the market, then ask your users. This point really boils down to the following: under the current state of AI, you can’t be complacent. Always look for ways to enhance your product and please your customer base. Continue to test, continue to build, then test again.

I’m curious to hear your thoughts on each of these areas. To you, what does starting and scaling a startup look like in 2026? What are other ways to defend your idea?

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

However, when you used to do a Google Search, you would get a series of sponsored ads related to the search. These would be at the top of your SERP - with some text indicating the result was in fact sponsored. For example, if you were to search “flowers near me”, then sponsored ads would sit at the very top of the page: same title, same URL, same grey description.

The only differentiated factor was a small bold “Ad” label in a tiny box next to the URL, easy to miss by design. There were typically 2-4 of these before the first organic result. For advertisers, that mean clicks. For consumers, it meant often visiting sponsored websites without even realizing it.

Subscribe now

AI-Integrated Search Changes Things

With Google’s AI platform Gemini integrating directly into Google’s search engine, I began to wonder… how is it going to work? The goal of an AI is to provide the most accurate information it can, and by ineffectively integrating ads into the platform, it could hinder this. So why not ask Gemini to share its take on the matter.

Turns out, these sponsored ads are now towards the bottom - unlike what was previously the case, where sponsored ads were located towards the very top. This raises the question, will that change? With the significant investments in artificial intelligence these major tech companies are making, will harder-to-find sponsored ads towards the bottom of the page provide enough of a payoff?

Image Credit: Searchengineland.com

Early Indicators of a More Dramatic Change

This seems like just a small part of the direction I see us heading with artificial intelligence. Right now, we are at the peak of increasing engagement and adding users (at least in the United States). However, we could soon be headed in a different direction: now that you use it, and now that you like it, now we have to start giving our investors a return on their capital.

ChatGPT recently began rolling out ads; Gemini and Perplexity have already implemented advertisements; and Meta AI plans to do so in February (with Claude as the ad-free outlier, solely focusing on subscriptions). What started as a “last resort” for these platforms is now a reality.

The question is, what extent will these advertisements influence responses? What data/privacy concerns might come of it? How will advertisements impact the churn rates and customer lifetime value of users? Only time will tell, but if I had to guess, we’ll learn a lot very quickly.

Where Do We Go Next?

Television introduced ads into the viewership experience via commercials. The Internet brought digital advertisements directly to your fingertips. The rise of social media influencers blurred the lines between advertising and purchasing via platforms like TikTok Shop and Instagram Reels Product Tagging. Who knows how ads will embed themselves into the user experience across AI applications.

As LLMs continue to become more powerful, more intuitive, and easier to integrate within legacy systems, one thing’s for sure: there’s going to be a need to pay for it.

Where do you see the industry headed next? What proprietary business models might we see as AI integration rates continue to skyrocket?

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment

Greater market opportunity means a much higher likelihood of business creation. Without a doubt, companies raising seed rounds today will inspire the next wave of entrepreneurs — particularly in an era where developing a deep understanding of a market, conducting competitive due diligence, and prototyping a product can be accomplished in a matter of minutes.

AI has given anyone with an internet connection access to vast amounts of knowledge in a matter of seconds. Greater feasibility means more builders discovering and developing new ways to solve the world’s toughest challenges. There might not be a more exciting time in history for both former and aspiring founders.

Subscribe now

Bagel Store Experience

The other day, I was at my hometown bagel store, which happened to be flooded with customers. I watched the owner — with whom I’ve become friends — running from one end of the store to the other, taking orders, checking finished ones, and greeting customers by name. Inching my way to the front, I couldn’t help but comment on how crowded his store was.

Seconds later, the owner pulled me aside. Pointing to an old picture of a small bagel shop, he told me that this was what his store looked like 30 years ago. A quarter of the size, a quarter of the staff, and 10% of the customers. He explained how tough it was to bring people through that door — but it never stopped him.

This brief interaction reminded me of the true meaning of entrepreneurship, and a lesson I encourage all founders and prospective ones to carry with them as they turn their ideas into action: AI might help it grow faster and make it easier to start and scale, but only passion can carry you all the way forward. We cannot let efficiency and automation cause us to lose sight of what matters most.

AI might build the runway quicker. But without passion, there won’t even be a plane.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable: https://bryan-hanley.lovable.app/

Leave a comment

But to solely look in these areas is a nearsighted approach to a long-term opportunity.

AI will impact countless industries. What we can’t forget to ask is not just how the industries currently touched by AI will be impacted — but how those industry changes will impact others. That second-order question is where I focus.

Subscribe now

I call this the Second Layer approach to venture investing:

Step 1a. What industries are most dominant today?

Step 1b. What industries are growing the fastest?

Step 2a. What opportunities might 1a and 1b lead to?

Step 2b. What risks might 1a and 1b lead to?

Step 3a. What solutions supplement the growth of 2a? → these are investments

Step 3b. What solutions mitigate the risks of 2b? → these are investments

Why this matters now

2025 was a top-heavy year — over 50% of all VC deal value came from just 0.05% of deals, led by decacorns like Databricks, OpenAI, SpaceX, and Anthropic. 2026 is expected to bring greater deal volume and less concentration as investors shift focus toward measurable outcomes over pure experimentation.

As liquidity returns to LPs — who are finally seeing returns as AI companies pursue enterprise partnerships, subscription tiers, and monetisation at scale — there is more dry powder available and greater openness to diversification.

The broader market remains focused on the first layer, but investors who back founders building for the second layer enter a less crowded yet highly consequential landscape”

The Second Layer is not compliance tech investing. It is a sourcing logic — a systematic way of finding the companies that dominant trends make inevitable, before the broader market recognizes them as a category.

I’m excited to share more thoughts about this perspective, and always welcome the chance to discuss — or better yet, debate it.

Relevant Links

Email: bryanhanleyvc@gmail.com

LinkedIn: https://www.linkedin.com/in/bryan-stanley-hanley/

Lovable Website: https://bryan-hanley.lovable.app/

Leave a comment