I don’t know who you are. Maybe you’ve seen me on Twitter or LinkedIn. Maybe we’ve never crossed paths. Either way, I'm writing this as a friend who cares about the version of you that wants to build something big (or not, both are good goals to have), maximise your potential, and live a full life.

I wish someone had sat me down in my late teens and said this out loud.

I’ve been thinking about it a lot lately: How to live my best life without dumping my health in a bin - staying as ambitious as I am, without treating my body like disposable fuel.

I’m sure you’ve seen that tweet that floats around every 2-3 months - “Hard work is overrated”? A lot of people read it as Gen Z being lazy. I don’t think that’s what’s happening. I think our generation is starting to notice something the hustle videos never mentioned: you can value your work and value yourself at the same time. You can love what you do without letting it quietly erode the rest of you.

I used to believe the opposite.

I started working at 17 out of ambition to do my best work, stubbornness of being financially independent and to support my parents. This meant classes, work, more work. No real life outside of it. I lived in one of the most liveliest cities in India, Gurgaon during my engineering days and I don’t remember stepping out of my college campus for fun, even once. I watched the “go big or go home” clips (please don’t judge me lol) and took them as gospel.

Work felt so joyful that nothing seemed to tire me (and it still does, probably will always do). Very early in life, it became my entire identity and took me years to realise what it can do to you. If a client didn’t close or a project stalled, the unhappiness didn’t stay at the desk - it followed me home, into my relationships, into my head.

Then the early warning lights started flashing. Migraines. Anxiety attacks. An autoimmune disease. The kind of health scares that make you sit up at 2 a.m. and ask, “What am I actually doing?”

I’m still a workaholic. Work still lights me up. But I’m deliberately building other things that can define me too, because when work is the only thing, the highs are higher… and the crashes hit everything.

For the nerds, don’t take my word for it: [recent studies by Indeed and Deloitte show up to 58% of Gen Z professionals are experiencing burnout - the highest rate of any generation.] Chronic stress doesn’t just make you tired; it messes with hormones and immunity, showing up as migraines, autoimmune flares, or that constant low-grade exhaustion so many women in their 20s are quietly living with (as a woman in 20s, trust me, it’s real).

What we brush off as “paying our dues” is actually laying the groundwork for irreversible damage. Relentless cortisol drives chronic inflammation, [creating the exact cellular environment where serious diseases like cancer thrive, according to research in ‘Brain, Behavior, and Immunity.’] You can’t out-earn structural damage to your heart and cells. [The WHO explicitly links working 55+ hours a week to a 35% higher risk of stroke and a 17% higher risk of fatal heart disease.]

We assume we can undo the physical toll once we have time and money, but..[the long-term CARDIA study proved that arterial damage done in your 20s is cumulative - the increased risk for heart attacks persists even if you adopt a perfectly healthy lifestyle by your late thirties.] The workaholic habits we normalise now are no badge of honor - they are quietly wiring our bodies for severe health crises we won’t be able to reverse in our 40s. [Credit to Gemini for populating the last 2 paras with scientific research]

Aaand, the good news? You don’t need to become Arnold Schwarzenegger in an instant because I know that sounds like a mountain to climb. But trust me, you don’t need a coach or a fancy routine if you can’t afford one. It starts small, and the fact that you’re even thinking about this means you’ve already done the hardest part. By starting small, I don’t mean strap a health band on your wrist while you eat trash every day, but instead, look inwards and start with some of these:

• 30 minutes of dumbbells + mobility workout a few times a week
  (go to any decently viewed YT video or check this app on iOS calls ‘Bend’)

• Make conscious choice of what AI tools to use and for what
  (for example using dictation tools can hurt cognitive health, there’s research showing writing actually reduces dementia risk - not just helps memory generally. My view: rely on dictation 100% and drop writing altogether, and you’re raising that risk, it’s also why writing is often prescribed to older adults to protect memory)

• Learning what a balanced plate actually looks like at home
  (read how to add protein to your fav dishes, why fibre is important, and so on)

• Get a full blood panel to know what you’re deficient in, and start fixing your basics

• Stand up from your desk for 15-20 min walks every 2 hours, learn exercises you do sitting in a chair

• Getting honest about your sleep
  (most of us need more than we admit)

• Look inwards, observe negative patterns (both, in action and thought), reflect on what emotional work do you need to do to reach the best version of yourself and sign up for mental health therapy

• Checking your screen time and noticing the dopamine loops big tech is very good at designing
  (ditch apps that get you into the habit of doomscrolling)

While we’re here - a reminder that endless scroll isn’t normal. I can go on and on about this so I’ll keep it short. Big tech is very invested in keeping us in those dopamine loops [using psychological tactics extensively documented by organisations like the Center for Humane Technology] so reflect on your behaviour and avoid it.

If you’re wondering, what can these small changes even help you get to, here’s how life looked like before vs. how it looks like now for me:

If you’re still thinking, it’s not my time, yet - Arianna Huffington learned this the hard way too [detailed in her book, Thrive] - she collapsed from exhaustion at her desk, broke her cheekbone, and later called the idea that success requires burning out a “collective delusion.”

Sure, some nights will still be late. That’s okay. The question is whether the other days are actually recovering you - whether you’re creating any real balance at all.

Ask yourself - For the body that’s doing so much for me, what am I doing for it?

Look around at the people who’ve been grinding for decades. There’s a reason so many of them suddenly become health freaks in their 40s and 50s. So, why wait to give yourself more years of the best version of yourself? You ONLY function at your best when your mental health, physical health, and inner self are in sync.

Remember, you don’t have to overhaul everything today. Just start somewhere. Notice your body. Protect a little more of your energy. Let work be something you love without letting it be the only thing that gets to define you.

I’m rooting for the version of you that gets to keep going - ambitious, joyful, and still standing strong years from now.

With hugs,
Someone who’s still figuring it out in real time

I keep wondering if this is who we were meant to be. Not the scrolling itself, but the way the scrolling has rearranged what happens inside my head before I’ve even said good morning to anyone real.

I notice I’m constantly adjusting - how I talk in meetings, what I choose to post, even the clothes I reach for. It’s not that I want to perform badly, in fact I’m the opposite - I do not care about anyone’s opinions, and yet, I’ve become who I am not. It’s that the performance has started to feel like the default setting.

I catch myself choosing what to write based on what might travel further, not just what I actually think. I notice the job I’m doing feels slightly different because of how it will look when I eventually post about it. Even the books I pick up or the places I want to visit get filtered through what they’ll look like on a screen. It’s not always conscious. That’s what makes it harder to notice. The algorithm doesn’t need me to try; it just needs me to keep feeding it the version of life that gets rewarded.

And now AI is sliding into the same current. It can write the post, generate the image, summarise the meeting, even suggest the “authentic” reply. The question isn’t whether the tool is good or bad. The question is what happens to the person who no longer has to sit with the blank page, or the awkward silence, or the thought that doesn’t come pre-packaged for an audience.

If the machine can handle the thinking and the shaping and the showing, what’s left for the part of us that used to do those things because they mattered to us privately?

I don’t think the answer is to delete everything and move to a cabin. Most of us aren’t going to do that, and we probably shouldn’t. But I keep coming back to a simpler test: when was the last time I did something - really did it - without any part of my brain calculating how it might land with someone else? Not to post it later, not to feel seen, not to get the little hit of validation. Just because the doing itself felt necessary or interesting or alive in that moment.

The attention economy isn’t evil in some cartoon way. It’s just single-minded. It only knows one direction: more attention, more signals, more engagement. Everything else - boredom, private thought, slow conversations, work that no one claps for - gets quietly deprioritised. Over time that single vector starts to shape what feels normal. We begin to treat our own inner life like content that needs an audience to be worth having. And slowly the difference between “what I actually believe” and “what performs like something I would believe” gets harder to feel.

I don’t have a neat solution. I’m still figuring out how to live inside this same system without letting it finish the sentence for me. But I know the cost of forgetting is real. When we lose the habit of checking in with the version of ourselves that exists when no one is watching, we also lose the raw material that makes us interesting to watch in the first place. The opinions that aren’t optimised, the tastes that don’t trend, the work that doesn’t photograph well - those are the parts that actually belong to us, the parts that make us human. Everything else is just renting space in someone else’s feed.

Maybe the real work isn’t fighting the machines or the platforms. Maybe it’s smaller and quieter than that. It’s noticing, once in a while, when I’m about to do or say or post something and asking, very simply: would I still do this if the like button disappeared tomorrow? Because the world we’re building will only be as human as we remember how to be.

I’ve moved from writing to speaking - dictating notes, memos, ideas - because typing feels like friction now. My repetitive tasks are handled by AI. I summarise podcasts and essays instead of listening to or reading them myself. I build small apps and websites for things I used to solve with a spreadsheet or a conversation.

Last week, I came across a word I didn’t know. I didn’t look it up. I texted Claude. My husband recently asked me to help him with something - I built an app to solve it. As an investor, I got tired of making decks, so I stopped making each slide from scratch.

I’m becoming lazy. In simple words. (And if you thought - hey, no - that sounds productive, seems like you know how to use AI well, hold that thought)

To put words in my perspective, I’ve basically turned myself into a machine that fuels my ADHD-like symptoms, a society-set metric that says more AI is good, faster learning is good, ship faster, think less, output more, and set myself on to a path that declines my social, emotional, and cognitive abilities.

Subscribe now

The moment it clicked

Two things happened that made me stop and look at what I was actually doing, and why.

The first was my husband, Rajat. As a corporate lawyer should, he keeps questioning me on everything I do with AI. Not in a dismissive way - more like a persistent, annoying audit of my daily habits. “Did you actually think through that, or did Claude think through it?” It’s the kind of question you wave away the first ten times and then can’t stop hearing.

The second was Michael Brenner - professor of Applied Mathematics and Applied Physics at Harvard and a research scientist at Google. He recently changed how he evaluates students in his class. His advanced applied maths class used to have problem sets that were genuinely hard. Then AI started solving all of them. So he flipped the assignment: students now have to come up with ten problem sets that AI cannot solve, verify the solutions themselves, and then defend one of those problems in an oral exam.

His reasoning was simple: because we have AI, students should do more. They should solve harder problems. They should learn more.

I went on a bit of an existential trip after that. I audited my own AI usage, just honestly asking myself: in the last month, when did I use AI to genuinely raise the bar on my thinking, and when did I use it to skip the thinking entirely?

The honest answer: sometimes I was probing, questioning, pushing my work further. But most of the time, I was outsourcing.

AI is not making us better learners or smarter, there’s data for this now

A study from MIT’s Media Lab tracked students writing essays over four months, split into three groups: one using only ChatGPT, one using search engines, and one using no tools at all. They measured brain activity with EEG scans throughout.

The group that used ChatGPT exclusively showed the lowest neural activation - and it got worse over time. Their brain connectivity in regions associated with critical thinking actually decreased across sessions. And 83% of the ChatGPT group couldn’t recall key points from their own essays when asked afterwards.

The researchers called it “cognitive debt” - the cumulative cost of letting a machine do your thinking. (Source: MIT Media Lab, “Your Brain on ChatGPT,” 2025)

A separate study from Microsoft Research and Carnegie Mellon surveyed 319 knowledge workers across 936 AI-assisted tasks.
The finding that stopped me: the more people trusted AI’s output, the less cognitive effort they applied. For 40% of their tasks, workers reported using no critical thinking whatsoever.

The researchers described it as a shift from doing to overseeing - and warned that it leaves people “atrophied and unprepared” when they actually need to think independently. (Source: Lee et al., “The Impact of Generative AI on Critical Thinking,” CHI ‘25)

Then I listened to a podcast episode with Brenner and two other Harvard researchers - Tina Grotzer, a member of the faculty of education, and a cognitive scientist who studies how people learn, and Ying Xu, assistant professor of education, who researches AI’s effect on children’s development. Their data, and insights hit differently because it was about learners. (If you want to read in detail, here’s the link to transcript in the Harvard Gazette)

A survey of 7,000 high school students found that nearly half felt they relied on AI too much. Over 40% said they’d tried to cut back and failed. Teenagers describing what is essentially an addiction pattern. I’m sitting here, an adult professional, recognising the exact same thing in myself - I tried to stop using Claude for every small question and lasted about three days.

The same survey found a significant drop in motivation to learn Maths and English after AI was introduced. As an investor, it made me question - Why should I invest time in research if AI can do it for me?

The researchers called it a “crisis of purpose.” If AI can do the thing, why should I learn the thing?

And there’s a creativity angle too. A 2024 study published in Science Advances found that writers who used generative AI produced stories rated as more creative individually - but collectively, the AI-assisted stories were more similar to each other. Everyone’s output got a little better, but everyone started sounding the same.

The researchers called it a social dilemma: individually better off, collectively losing novelty. (Source: Doshi & Hauser, Science Advances, 2024)

That last one bothers me more than the rest.

Because the whole pitch of AI is that it’s a thinking partner. But if everyone’s thinking partner is drawing from the same statistical well, what happens to the range of human thought?

What conscious use of AI actually looks like

So what do you do? You can’t un-know AI exists. And honestly, avoiding it entirely would be its own kind of stupidity. I’ve started making changes. They’re small, and I’m not consistent, but they’re deliberate.

The workflow I’m optimising for now is metacognition - which is really just a fancy word for knowing whether you’re actually thinking or just going through the motions. Can I explain this output without looking at it? Did I have a point of view before I prompted? Do I understand why the AI’s answer is right, or am I just trusting that it is? What’s AI missing?

Here’s a glimpse of what that looks like in practice for me:

1. The order you do things matters more than the AI tools you use.

If AI is step one, you’re outsourcing. If AI is step three, you’re augmenting.

And I’m not talking about work like summarise unread emails for me, I’m talking about work that requires expansive thinking and knowledge. For example, when I’m writing an investment memo, I spend the most time IRL, gathering intel from people - operators, experts, founders - and then I use my own thinking to form insights.

AI should be your last step, to probe, to add, to pressure-test. And if it’s not, your ability to analyse will start to dull.

2. Protect the inputs. Don’t just optimise the outputs.

I stopped dumping every podcast, essay into NotebookLM. I’ve started listening to the longer ones end to end - actually listening - and taking notes by hand in a diary or Google Keep. For really long research papers, I still use NotebookLM because I’m a visual learner, but I sit with the summary afterwards. I debate it with friends, with my husband, I bring it up on social media.

I’ve gone back to reading newspapers (reading offline also leads to better comprehension and memory retention). I take photos of articles that stand out - send them to friends, discuss, debate. Sometimes I send that photo to AI to give me more data around it, brainstorm. But the thinking happens before the prompting, not after.

I’ve been using an AI tool called Oboe to learn new topics. What I love about it is - it engages you with questions, great visuals, and has exceptional contextual depth. It makes you work for the understanding.

Even with writing, I’m trying to be more conscious about when I speak versus when I type, and exploring workflows that inspire me to creatively think. I use Wispr for dictation, and it’s great for certain things - if I have a complex thought and I need to get it out before I lose it, speaking is faster and I’d get lazy if I had to write it all down. That’s how this blog started, honestly.

But I’ve noticed that when I need to actually structure my thinking - organise an argument - writing by hand or typing it out myself does something different to my brain. And diary, because writing forces you to commit to a sequence, and that’s where comprehension and memory actually form. (worthy shout-out to spiral AI, I use Claude now for this but it’s really good at helping you put a structure to your thoughts)

And this reminds me of my conversation with my X friend, Kuldeep that I must highlight as well:

3. Use AI as a sparring partner, not a ghostwriter.

Going back to the Harvard podcast expisode I was talking about, Grotzer talked about teaching students metacognition by having them build Venn diagrams of what their mind can do versus what AI can do. It sounds simple. It’s not. (In fact, if you’re reading this, try this exercise yourself)

When you actually sit down and ask yourself, “what am I better at than a language model?”, you start to understand where to protect your own cognition and where delegation is fine.

Brenner’s model is the one I keep coming back to as well: don’t ban AI. Raise the bar so high that AI becomes a floor, not a ceiling. And treat guardrails not as punishment but as protection - not for today’s convenience, but for your future cognitive ability (bye bye, dementia).

The uncomfortable truth, and the podcast researchers said this clearly, is that self-regulation with AI is a skill most of us haven’t developed. AI tools are built to be frictionless, accommodating, always ready. They have everything they need to make us dependent.

I heard a cognitive scientist say this somewhere that stayed with me too - that AI cannot make you smarter. If you’re already thinking clearly, great, you’ll get useful outputs. But if you’re not, your prompts will be bad and your results will be worse. AI is an amplifier, not an educator. And most of us are not prompting well. I’m including myself.

And I strongly believe, the people who’ll come out of this well - the ones I’d actually call intelligent knowledge workers - are the ones who will know where to slow down on purpose. Not because they’re anti-AI. But because they know that if you outsource the thinking, you’re not raising the bar on anything. You’re just getting faster at being average.

The wage premium illusion

And if you’re thinking, hey I need to use AI every day no matter what, I don’t care about climate or cognition declining, but hey at least it’ll help me get a good bonus/get better jobs, you’re in for a surprise.

According to PwC’s 2025 Global AI Jobs Barometer - which analysed nearly a billion job postings across six continents - workers with AI skills commanded a 56% wage premium in 2024, double the 25% from the year before. (Source: PwC, 2025 Global AI Jobs Barometer)

Sounds incredible. But that premium doesn't go to everyone who uses AI - it goes to people who can use it well. And using it well means you already need to know what good looks like in your field. You need to know what to ask, how to spot when the answer is wrong, and when to override the output entirely.

And here’s the part PwC’s own data reveals: between 2019 and 2024, US job postings in occupations most exposed to AI grew by just 1% per year. Meanwhile, roles least exposed to AI - bricklayers, food preparers - saw 20% growth. AI isn’t destroying jobs. But it’s also not creating the flood of new ones that the enthusiasm might suggest. And I don’t even need to get pull stats on layoffs.

AI encodes human expertise into models that replicate it across millions of users without human intermediaries. The productivity gains flow up to companies and investors, not sideways to the workers whose expertise was encoded in the first place.

Where does this leave us?

I know the case for AI in drug discovery, climate modelling, personalised medicine. Real, credible, important.

But for consumer use cases (there are of course exceptions such as using AI for fixed workflows like astrology) - for you and me, using AI every day to think, learn, write, decide - I’m not sure it’s the force multiplier everyone says it is. Not until we learn how to use it consciously.

This reminds of an excerpt from Scott Barker’s (ex-VC, exploring the psychology of meaning) recent blog:

I read that and thought: I’m at the beginning of that curve. Most of us are. Except now the machine we’re turning ourselves into is being accelerated by another machine.

The questions I’m sitting with

If AI can’t make us smarter - only amplify what’s already there - what happens to a generation that outsources its thinking before it’s developed the thinking worth amplifying?

If teenagers (including adults) are already trying and failing to regulate their AI usage, what does that look like in ten years?

If the productivity gains go to companies and investors, not workers, and the wage premium only accrues to those who were already skilled - who is AI actually for?

If AI-assisted creative work is individually better but collectively more homogeneous, what happens to the range of ideas in a society where everyone uses the same tools?

Are we building towards AGI because it’s a worthy goal, or because the momentum is too lucrative to question?

And the one that keeps me up: for everyday human use, are we making ourselves dumber? Slowly, comfortably, one outsourced thought at a time?

I don’t know where this takes us. But I’m increasingly sure of one thing: AI alone is not the force multiplier we keep telling ourselves it is. The force multiplier is human thinking, augmented by AI. In that order. The thinking has to come first.

So, the next time you are peeking over the shoulders of somebody’s “thinkstack” of 20 tools, ask yourself this: why do you do what you do in the first place? when did faster and more efficient become more important than being honest, intelligent?

I love food - I take cooking classes everywhere I travel to and grew up on incredible home-cooked food, thanks to my mother and grandmothers.

Never learned to cook though. And nothing kills a teenage girl’s interest in the kitchen faster than being told it’s preparation for serving her future husband.

Years later, the day finally came. When I moved in with my husband in 2023, we made ourselves a promise: we’d handle household chores ourselves. No external help.

One of our major tasks was figuring out cooking. We’d hired several cooks before this, but it never worked flawlessly. Either they couldn’t cook the dishes we liked, or the taste was off. Through no fault of their own, they had to service multiple houses to earn a basic income - which meant they could only spend 30-40 minutes max in our home. High flame, excess oil, you know the works. So, we took it upon ourselves to figure this out. Step one: getting my mom over for 10-15 days to teach us the basics.

Then I started searching for products that could make cooking feel less like a chore. I’d heard of semi-automatic devices like Thermomix in the West (fun-fact: they clocked $1.8B+ in revenue in 2024). I’d seen Instant Pot conquering American kitchens and some newer Indian ones. But when I actually researched what would work for us - for Indian cooking, in an Indian kitchen - I kept hitting dead ends.

Products that worked for pasta didn’t handle tadka well. Devices built for Western kitchens assumed you had huge counter space. The “smart” appliances mostly just had Bluetooth connectivity and recipe apps, but couldn’t actually reduce the mental or physical load of cooking.

Today, many urban homes are full of technology - Google Home, Alexa, motion-sensor lights. But our kitchens? Walking into mine as a 26-year-old woman, I had no real tech assistance. (And I’m not talking about peripheral help like microwaves or air fryers - I mean something that actually changes how you cook daily.) The experience left me completely frustrated.

To no one's surprise, I found that Indian households spend 13-14 hours per week cooking, while the global average is 6 hours (GfK Global Study, 2019).

What’s even more frustrating is that breaking this down by top cities in India reveals the burden is even heavier for women: Delhi leads at 3.4 hours per day, followed by Kolkata at 2.9 hours, Chennai at 2.8 hours, and Bengaluru at 2.7 hours (Open City, “Childcare and Domestic Work - Time Use Survey,” 2025.)

So, why is India’s cooking time more than double the global average?

• Technical complexity - Indian cooking’s technical complexity lies in its reliance on water-based methods that demand precision and intuition. Unlike roasting or grilling where heat does the work, techniques like tempering spices in exact sequences, controlling the reduction of gravies to achieve specific consistencies, and layering flavours through staged additions require deep understanding of timing and chemistry.
  The same curry can have 15+ spices added at different stages - some bloomed in oil, others simmered in liquid, some finished raw - each step transforming the flavour profile. Most critically, these techniques are rarely standardised - they’re learned through observation and muscle memory, making them nearly impossible to replicate without years of practice (like many say, a superpower only our moms have).

• Cultural expectations of freshness - Indian households predominantly cook fresh meals daily, often twice daily. On the other hand, western food culture has normalised meal prep, leftovers spanning 2-3 days, and frozen meals as acceptable alternatives. Our habits of cooking and eating fresh food primarily comes from our warm, humid climate that couldn’t sustain food storage for long without refrigeration. In fact, Ayurvedic texts which is followed by many in India, classify stale or “tamasic” food as undesirable because it’s believed to reduce vitality and affect physical and mental health.

• Patriarchal household structures - In India, it has unfortunately always been a women’s job to cook and deliver good food on the table, never men’s. It is expected, unpaid, and often unassisted. Women bear the full 13.5 hours every week largely alone.

  The gendered nature of cooking labour also meant the problem stayed invisible to those who (read: men) traditionally controlled household purchasing decisions. We got money pouring into R&D for electric razors, power tools, and golf simulators while women to date cook the same way their great-grandmothers did.

• Fewer acceptable convenience alternatives: Western markets have decades of R&D in high-quality ready-meals, meal kits, and semi-prepared ingredients that taste good enough to substitute for home cooking. India’s ready-meal market has historically suffered from the “doesn’t taste like home-cooked” problem that severely limits repeat purchase.

We have AI that can beat grandmasters at chess, but we're telling Indian women to spend 13.5 hours a week at the stove? The math isn't math-ing. If technology can replicate judgment well enough to approve loans and recommend surgery, it can sure as hell learn to cook Indian dishes. As a VC and a woman watching this gap, I had to dig in - equal parts rage and curiosity.

I explore four questions in the blog:

1. Why did 40 years of kitchen innovation fail to reduce the 13.5-hour cooking burden in India?

2. Why does this wave succeed where previous attempts failed?

3. What does the next 5-10 year trajectory look like?

4. What should investors and founders in India watch for?

Let’s start with the first question.

I. Four decades of failed convenience

For 40 years, consumer demand for kitchen convenience has been evident. Brands like Maggie demonstrated it and the growth of convenience food (ready-to-cook + ready-to-eat) proved it. According to Redseer's report 'The Ready-to-Eat Opportunity: Decoding India's Convenience × Health Pivot', India's addressable consumer base for convenient foods has grown from 50 million to nearly 80 million in just two years. Market size is nearly $6B, projected to hit $11.3B by 20233.

The consumer need in food convenience is growing, there’s a behaviour shift taking place in how consumers eat food. So why hasn’t tech entered the chat? Let’s first rewind to the days when Maggi entered India and understand where convenience in food exists today.

Maggi’s success shows demand for convenience - at the edges

Maggi arrived in India in 1982, positioning itself as “2 minutes to cook” convenience. The product found product-market fit and maintained it for decades.

Looking at Nestlé India’s financials tells the story of sustained demand: Maggi represented roughly 23% of total revenue in 2010 (around ₹1,400 crore / ~$280 million), grew to 26% by 2015 (₹2,100 crore / ~$420 million), and by 2023 had reached an estimated 30% (₹6,100 crore / ~$730 million out of ₹20,100 crore total). By 2024, the Maggi portfolio had grown to an estimated ₹7,200 crore / ~$860 million (Finception 2018; OutlookBusiness 2025; buildd.co; Business Standard 2024).

Important context: “Maggi” as a portfolio includes instant noodles, sauces, masalas, and other products. The instant noodles segment alone is estimated at 60-65% of this revenue, but Nestlé doesn’t break out exact figures publicly.

For the kids that missed the golden 90s era, here’s a Maggi ad from Doordarshan (Govt. broadcaster) in the 90’s with their 2-minute tagline.

This demonstrates sustained consumer willingness to pay for convenience. But the product’s positioning also reveals its limitations. In urban India, Maggi only succeeded as: late-night snacks, emergency meals when cooking isn’t possible, quick fixes for children or singles, weekend laziness solutions.

What Maggi didn’t replace for urban India: planned family dinners. The daily dal-roti-sabzi. Sunday biryani. The meals that make up the bulk of those 13.5 hours.

Every innovation since has followed a similar pattern. Quick commerce (Swiggy Instamart, Zepto, Blinkit) delivers ingredients faster - but doesn’t reduce chopping, stirring, or monitoring time. Ready-to-eat and restaurant meals promise convenience but suffer from the “doesn’t taste like home” and often hygiene problems that limits repeat purchase.

However, a new category of pre-prepared ingredients is emerging that sits between raw ingredients and ready-meals: pre-cooked staples without preservatives, that reduce prep time without sacrificing the “home-cooked” experience. One of the companies I closely work with at Antler called Freshcon sells pre-cooked rajma, chana, sweet corn, and soon rice. They’re seeing surprising traction - crossed 1M+ customers across 1600+ cities in India in under 12 months. This trend suggests consumers are willing to outsource the time-intensive prep work (soaking rajma overnight for 6-8 hours, pressure cooking chana for 45 minutes).

These complementary products are creating an ecosystem where the remaining bottleneck becomes increasingly obvious: the actual cooking process itself. If I can get ingredients delivered in 10 minutes via Blinkit (I can even order peeled garlic), buy pre-cooked rajma to save prep time, and have a recipe on my phone, the last friction points are:

• The mental load of getting the recipe right - recalculating measurements, remembering what to do when, and researching through multiple recipes to find the most trusted one

• The physical load of standing at the stove for 30-45 minutes stirring, monitoring, and adjusting. This is precisely the gap that cooking tech/AI can address.

So, why didn’t we have enough companies build products that attacked this core time sink? Three structural barriers explain the gap.

a) The stakes were too high

Alexa misunderstands your song request? You say “Alexa, play something else.” Minor friction, minor consequence. You get the wrong order from an e-comm website, you ask for a return, job done.

If a cooking appliance burns dinner? You’ve wasted an hour of prep, ₹500 (~$6 USD, and that’s not a small amount for most India) in ingredients, your mood is spoilt, and now your family orders takeout. The failure isn’t inconvenient. It creates cascading problems: wasted time, wasted money, hunger, family frustration, loss of trust in the product.

Cooking demands much higher reliability because the stakes of failure are high and so is the complexity of getting it right.

Previous automation attempts couldn’t clear this reliability bar. Pre-AI, cooking tech relied on:

• Hardcoded temperature and time sequences (no adaptation to ingredient variance)

• Binary sensors (is water boiling: yes/no, rather than “how vigorously” or “needs 30 seconds more”)

• Recipe following without understanding (if user substitutes ghee for oil, system doesn’t adjust)

• No feedback loops (if curry is too thick, system doesn’t recognise or correct)

These systems worked acceptably for simple, standardised tasks (rice cooking, pressure cooking). They failed for the complex, judgment-based tasks that consume most cooking time, which results in average taste (a north star metric for cooking tech companies besides reducing mental/physical load and time).

The tech gap earlier wasn’t just AI capability. It was also technology (can you detect curry consistency, oil separation, dal texture reliably and cheaply?), edge computing (can you run adaptive models on device-level hardware?), thermal control precision (can you maintain and adjust temperature zones dynamically?), user interface design (can you make it simple enough that setup doesn’t require engineering degree?). More on this in the following subsections.

But let’s say we crack the tech after all these years - who’s going to buy it? Earlier, we covered how patriarchy is precisely why no one bothered investing in making cooking easier for women; it simply wasn’t worth solving for. So, what’s changed?

b) The time nobody valued (until now)

The 13.5 hours of cooking fall predominantly on women. This is the likely biggest structural reason the problem stayed unsolved.

In patriarchal household structures - which still dominate Indian homes - women’s labour in the kitchen is “expected.” Not valued as economic activity. Not compensated. Not even fully visible as work.

This creates a purchase decision problem. The people spending 13.5 hours (women) often lacked purchasing power and decision authority. The people controlling household spending (often men) didn’t experience the problem directly, so saw nothing urgent to solve. The kitchen “works” - meaning food appears on schedule. The hidden cost (women’s time and energy) wasn’t factored into the household’s economic calculations.

This dynamic still exists in many households. But the economics are shifting. Fast.

Dual-income families are turning those 13.5 hours into an opportunity cost with an actual, measurable price tag. Every hour in the kitchen is an hour not spent on career advancement, additional income, skill development, or rest. When women started earning salaries, their time gained economic visibility.

The most widely cited statistic on dual-income shifts comes from Morgan Stanley’s 2024 report. In rural India, dual-income households rose from 22% in 2018 to 42% in 2024 (Morgan Stanley 2024). For urban areas specifically, growth has been steady but significant. In 2011, 26% of urban households were dual-income (Census 2011). By 2024, estimates suggest roughly 30-35% (some sources even claim 40%), with urban female labour force participation rising to 27.2% in 2024 from approximately 20% in 2019 (Bain India; MoSPI 2024).

Women’s economic participation is also accelerating across every metric:

• Bank account ownership: 43% of women had bank accounts in 2014, growing to 78.6% in 2021 (World Bank Global Findex). As of 2024, women hold 39.2% of all bank accounts in India (MoSPI 2024). The gender gap in account ownership is closing faster than we think.

• Credit access: Women seeking credit grew at 22% CAGR versus men at 11%. Active women borrowers reached 83 million in 2024 (TransUnion CIBIL; NITI Aayog).

• Investment accounts: Women holding demat accounts jumped from 6.7 million in 2021 to 27.7 million in 2024 - a 4x increase in three years. Women now account for nearly 1 in 4 new investors (MoSPI 2024; SBI Research).

We’re still away from the finish line, but these numbers represent a fundamental power shift in household economics. The person spending 13.5 hours in the kitchen now:

• Opens her own bank account

• Takes loans in her own name

• Makes investment decisions independently

• Controls purchase decisions for household goods

Women are also effectively the “gatekeepers” of household purchasing in India. Whilst men may often transact the final payment, women increasingly dictate the choice of brand and product. Women control or influence 85% of all consumer purchasing decisions in India (Fortune India; TechCrunch). Among salaried urban women, 98% are now actively involved in long-term family financial decisions like home loans, insurance, and education (DBS Bank-CRISIL Survey 2024).

Crucially, 47% of these women now make financial decisions independently - cutting out the male “approver” role entirely (DBS-CRISIL 2024).

I see this in my own extended family that’s spread across T-1 to T-4 towns of India - the decision making has switched from being men-only in the 2000s to 50-50 now in most cases.

The growth of e-commerce reinforces this shift. E-commerce in India is currently being driven more by women than men, particularly in tier 2 and tier 3 cities. 60% of women internet users prefer shopping online, compared to 52% of men (Home Credit India, “How India Borrows” Report 2024). Women now constitute 48% of all active online shoppers in India - nearly achieving gender parity (IPG Mediabrands; Brand Equity 2024). E-commerce platforms reported a 34% year-on-year increase in female shoppers in 2024, significantly outpacing male user growth (Indian Retailer; Flipkart Trends 2025).

The implications for kitchen tech are direct: the people experiencing the problem now have the economic power to solve it. When purchasing power shifts to the people bearing the time cost, demand for time-saving solutions becomes economically rational and actionable.

Innovation happens when the person with the problem gains the power to solve it. That power has shifted.

c) Some people actually want to cook

This was one of my biggest learnings.

Globally, 42% of home cooks report that concerns about ultra-processed foods have made them cook more at home compared to a year ago (Eastlake via FoodNavigator, 2023). India is catching up to this trend. Ultra-processed foods - instant noodles, packaged ready-to-eat meals, frozen dinners - are increasingly linked to obesity, metabolic syndrome, and cardiovascular issues.

Making meals healthier is consistently the top reason consumers cite for cooking from scratch. Beyond health, a significant portion simply enjoy cooking itself - the creative process, the satisfaction of making something, the ritual of preparing food for family.

But the data on what home cooks actually want reveals a nuanced picture: 73% prefer recipes with short ingredient lists, 64% want dishes that only require one pot or tray (Eastlake via FoodNavigator). They’re not asking to stop cooking. They’re asking for simpler cooking.

Consumer behaviour doesn’t shift overnight - it shifts in phases.

This reminds me of growing up with Weikfield custard and jelly mixes. I’d feel like a queen after making custard, watching jellies wobble perfectly on a plate once they’re set. It wasn’t “ready-to-eat” - it was “ready-to-make-easily/ready-to-cook,” and that made all the difference. I still got to make something. I just didn’t have to stress about whether it would set properly or taste right.

That’s exactly what happened at scale during the pandemic. Weikfield - a homegrown brand selling custard powder, jelly crystals, and cake mixes - saw 95% of its demand suddenly shift to consumers (up from just 20-25%), not restaurants. People weren’t ordering cakes from bakeries; they were making them at home using mixes that eliminated the guesswork. The company is now targeting to double revenue in three years (2021).

These aren’t people abandoning their kitchens. They were people who still wanted the pride of “I made this” - just without two hours of measuring or worrying about failure. The data and behaviour clearly signals that RTC will grow far more exponentially than the RTE segment for at least the next few years for the very same reason - people want to cook.

This is also the pattern cooking tech will follow: Cooking tech that enables people to cook simpler/faster will scale faster than tech that cooks for you.

My survey of 58 respondents reveals why. When asked which tasks they’d most want automated, 55% prioritised chopping and prep work as their #1 choice, not cooking itself. But the priority shifts dramatically by cooking frequency:

• Minimal cooks (0-3 hours/week): Split between chopping (39%) and cleanup (26%)

• Regular cooks (4-7 hours/week): Overwhelmingly prioritise chopping (60%)

• Heavy cooks (8+ hours/week): Want chopping plus temperature control and stirring - the judgment-based tasks requiring constant attention

As cooking frequency increases, the bottleneck shifts from “I don’t want to start” (cleanup deterrent) to “I can’t scale” (monitoring burden). This suggests different products for different segments: smart cleanup solutions for infrequent cooks, automation of judgment tasks for frequent cooks.

Meanwhile, qualitative responses on most-enjoyed cooking moments revealed consistent themes: “watching the masterpiece unfold,” “the whole process feels like therapy,” “the satisfaction of a dish coming out well,” “feeding family,” “when I taste it and it’s exactly what I wanted.”

These aren’t people who hate cooking. They’re people who hate standing at the stove monitoring dal for 30 minutes or spending 20 minutes chopping vegetables before they can start the enjoyable part.

Why do consumers want short ingredient lists and one-pot meals? Because these reduce the burdensome aspects: short lists mean less shopping complexity, fewer storage needs, lower cognitive load. One-pot cooking means less monitoring, less coordination, easier cleanup.

This explains why in cooking tech, full automation may see slower growth. Consumers don’t want their kitchen to become a vending machine - yet. They want it to become easier to use first. Technology that assists cooking will scale before technology that replaces it entirely. The emotional need for “I cooked dinner for my family” doesn’t disappear - it just evolves.

That shift creates an opening. The question is: what’s filling and enabling it?

II. The Equation Today is Changing Rapidly

Three factors converged in 2024-2025 that created a different outcome than previous automation attempts. Critically, these factors interact: each one enables and reinforces the others.

Technology has reached some reliability threshold

For decades, the problem was simple: cooking is wildly unpredictable. A tomato in Mumbai has different water content than one in Delhi. Rice from one harvest cooks differently than the next. Previous automation couldn’t handle this variability.

Today, these systems can manage such complexities in cooking and more. They make decisions, not just follow programmed sequences.

So what’s different today? I asked Prithvi (former Embedded Software Lead, Posha, now Co-founder & CTO, Stealth). He mentions, between 2022-2025, three major AI advances converged to make reliable cooking automation possible:

• Vision-Language Models (VLMs) made cooking AI practical: VLMs use the same transformer architecture that powers ChatGPT, but for understanding images and video. Earlier cooking AI would’ve required massive, manually labeled datasets (“this is 70% cooked onion”, “this is burnt garlic”). Modern VLMs learn from far less labelled data, understand context from natural descriptions, and generalise across cooking scenarios. What used to take months of data collection now takes weeks. And with optimised, lightweight versions of these models, real-time visual recognition (”is the oil separating?”) can now run on-device without cloud dependency.

• Multimodal AI matured: Instead of relying on a single input like a timer, systems now fuse multiple data streams like visual appearance, thermal readings, weight changes, even sound to understand cooking state.

• Edge AI hardware became affordable: Chips with built-in neural processing units used to add significant cost. Now, capable edge AI hardware has dropped to consumer-friendly price points, making smart cooking appliances economically viable for households.

And on the technology at large:

• Vision models have crossed a threshold: Modern vision models can now pick up minute visual cues like oil shimmer, colour changes, bubbling patterns, texture formation, things humans subconsciously use while cooking. Earlier systems relied heavily on timers because machines couldn’t “see” well enough. Today, vision adds real awareness.

• ⁠Sensor fusion instead of single signals: Instead of relying on just time, systems now combine multiple inputs, visual appearance, temperature, weight changes, and even sound. This gives a much more holistic understanding of what’s actually happening in the pan.

• Adaptive control replaces fixed recipes: Traditional automation follows rigid steps. Modern ML models continuously adjust heat, timing, and actions based on real-time feedback.

• Edge AI is finally practical: Modern SoCs come with dedicated Neural Processing Units (NPUs). This means heavy vision and decision models can run locally on the device, in real time, without cloud latency. Cooking decisions now happen in milliseconds.

• Massive reduction in data and annotation bottlenecks: Earlier AI approaches required enormous, manually annotated datasets. Vision-Language Models (VLMs) drastically reduce this burden, models can learn from far less labeled data, understand context better, and generalise faster. Training and iteration cycles are now weeks instead of months.

• Small Language Models personalise the experience: Compact, on-device language models can store user preferences, spice tolerance, doneness preferences, and adapt recipes dynamically. Cooking is no longer static or one-size-fits-all; it evolves with the user over time.

• Human-in-the-loop learning: Simple user corrections “cook it a bit longer,” “less oil next time” directly feed back into the system.

The AI inflection is real, but there’s also a deeper infrastructure problem that’s held back Indian cooking tech for decades. I spoke with Aditya (Founder, Cibos), who’s building DC heating coils and compressors domestically out of Ahmedabad, Gujarat. He says, “India’s cooking tech constraint isn’t just about software, it’s thermal infrastructure.”

Consider this: India imports 17,000+ shipments of basic heating elements annually from China, Germany, and Italy. Cooling coils, compressors for refrigeration come from China, Taiwan, and Malaysia. When your core thermal hardware, heating coils, heat exchangers, micro-compressors, comes from overseas, you can’t design for Indian conditions. You’re stuck designing around import availability, not user needs.

The real unlock happens when thermal infrastructure is built locally with design control. Then you can finally build appliances optimised for DC ecosystems (solar + battery), local energy realities (intermittent grid), and actual Indian cooking workflows, not just repackaged imported components with an AI layer on top. When these thermal building blocks are made locally, Indian innovators can break free from legacy supply constraints and build products tailored for how Indians actually cook, with the energy infrastructure we actually have.

Upliance AI’s September 2025 launch of Upliance 2.0 demonstrates how far the technology has come. According to the company, their latest device, it’s 40% faster than its predecessor at completing comparable recipes. It automates chopping, stirring, kneading, and sautéing through smart jars and integrated sensors. It ships with 750+ preset recipes that update over Wi-Fi, but more importantly, it allows customisation and learns from adjustments.

[Exclusive] Inside Upliance AI on learnings, user insights, and what’s next

To understand what’s actually happening on the ground beyond marketing claims, I spoke with two people with unique insider perspectives: Mahek, co-founder and CEO of Upliance AI, and Chaitanya, the former Head of Product at Upliance AI. Both were generous with internal data and candid about what’s working - and what’s still being figured out.

An important philosophical distinction

“Upliance doesn’t automate, it assists,” Chaitanya explains. “You still have to spend some time in the kitchen doing basic prep and adding ingredients during the cooking process.”

This differs fundamentally from fully automated approaches, which require more demanding prep - cutting vegetables into individual boxes, portioning masalas - but promise hands-off cooking once started.

Chaitanya breaks down the Indian market into segments:

1. People who enjoy cooking and dismiss new tech

2. People who don’t want to cook at all

3. People who want control over what they eat and would use tech

4. Tech enthusiasts who want automation

“Segments three and four are viable targets for cooking assistance,” he notes, “but today, this cohort is small, and needs to be expanded.”

Here’s the obvious question: why would households that can afford a ₹5,000-8,000/month cook spend ₹25-30K on a device?

The answer is that they solve different problems. A cook handles daily dal-roti, known recipes, prep work - solving “ don’t have time”, but not for consistent taste, diversity in recipes, and more. Cooking tech handles variety, experimentation, restaurant-style dishes you don’t know how to make - solving “I want to cook something different but don’t know how.”

“Some people equate it to a few month’s cook salary and find it easy to rationalise,” Chaitanya observes. But more importantly, these devices coexist with cooks, not replace them. There’s even an edge case Chaitanya mentioned: “a few users made their cook use the Upliance and now suddenly the cook could cook 500+ recipes.” Not replacing human labour, augmenting it.

Mahek’s answer surprised me: “Our biggest competitor is ordering food from Swiggy and Zomato. In most households we replace meals ordered from outside.” The device isn’t competing with the ₹5,000/month cook who does prep. It’s competing with ₹300-500 food delivery orders that add up to ₹12,000-20,000 monthly. The payback calculation suddenly makes much more sense.

The dual-income households in metros increasingly have cooks for daily basics but still want control over special meals, experimenting with new cuisines, or ensuring specific dietary needs. That’s the wedge.

What my survey reveals

My survey of 58 respondents (median age 28-35, predominantly metros, 70% dual-income households) shows nuanced patterns.

Among those cooking 4-7 hours/week:

• 60% prioritise chopping as their #1 automation need

• They’re comfortable cooking but want repetitive prep eliminated

• 21% of couples show willingness to pay ₹20,000+

Among those cooking 8-12+ hours/week:

• Want chopping plus temperature control and stirring - the judgment-based tasks requiring constant attention

• Only 12% willing to pay ₹20,000+, validating they need accessible pricing

• Cite “lack of variety/boredom” as a top friction when cooking 5+ cuisines monthly

By contrast, those cooking 0-3 hours/week split between chopping (39%) and cleanup (26%) - they’re deterred from starting at all. These are the “don’t want to cook” segment that full automation targets.

The pattern: As cooking frequency increases, the bottleneck shifts from “I don’t want to start” (cleanup deterrent) to “I can’t scale this complexity” (monitoring burden, variety management).

This validates that these aren’t just different price sensitivities - they’re genuinely different value propositions. Couples pay a premium for mental load reduction. Families need pure time efficiency at accessible prices.

Notably, 22% of Upliance sales currently come via EMI financing - a significant indicator that buyers view this as essential infrastructure, not just an impulse gadget purchase.

The psychological shift

“For most people,” Chaitanya observes, “the fact that a complex curry like paneer butter masala - which they had no idea how to make - was created by them just mindlessly following instructions and throwing in ingredients is a good first aha moment.”

This psychological shift from ‘I can’t make this’ to ‘I just made this’ is the conversion point where utility becomes satisfaction. It’s also evidence of AI democratising skill access - complex cooking techniques that previously required years of practice are now accessible through guided assistance.

Users liked that they could control the ingredients and make a “healthy” paneer butter masala at home - addressing both the aspiration (restaurant-quality dish) and the health concern (controlling oil, salt, quality of ingredients). There’s a sense of achievement in cooking yourself. Home-cooked food carries cultural virtue.

The cross-cultural opportunity

“What’s fascinating is the cross-cultural discovery happening,” Mahek notes. “We end up introducing a lot of people in South India to Pav Bhaji and Chole, while people in the North are cooking Palyas. Basically everyone is bored of their own staple, but luckily India has multiple cultures with multiple staples within it.”

This shows up in my survey data too: respondents cooking 5+ different cuisines monthly cite “lack of variety/boredom” as a top friction, whilst those cooking 1-2 cuisines cite “time/prep.”

The insight: the job-to-be-done isn’t replacing daily dal-roti (which people know how to make and often have cooks for). It’s enabling variety without the learning curve and time investment of mastering new cuisines.

“People who cook a couple of times a week on these machines see it as a ‘treat,’” Chaitanya adds, “and do things they wouldn’t cook by themselves otherwise. Restaurant style dishes like paneer curries, pulaos, pastas are very popular.”

The retention challenge

“There is a challenge of long-term retention - like most devices, how do you keep them coming back after the initial honeymoon period?” Chaitanya admits candidly. “And even if they enjoy the product a lot, sometimes they can ‘graduate’ to cooking on the stove. Or hire a cook later for some reason and never get back.”

This is the behaviour change risk that could prevent category formation. Perfect technology doesn’t guarantee sustained adoption.

But the positioning is smart: “The hardest challenge we can’t solve is to convert ‘Maa ka Khaana’ to code,” Mahek admits. “So we don’t try. Instead we focus on getting you to cook things you ate when you travelled or what you saw on social media.”

Don’t compete with emotional, nostalgic cooking (Maa’s version of dal, Dadi’s gajar halwa). Compete with the food people saw on Instagram last night, the restaurant dish they wish they could make at home, the cuisine they tried whilst travelling.

Who adopts first?

Based on the patterns above, two core segments are likely to adopt by 2030:

Segment A: Young working couples with no kids/DINKs (metros, ₹20L+ dual household income)

• Cook 4-7 hours/week, want variety and experimentation

• Well-travelled, Instagram-influenced, comfortable with tech

• Kitchen space is scarce; any device must earn its counter spot

• They enjoy cooking but hate repetitive prep and the mental load of “what’s for dinner”

Segment B: Nuclear families, especially working mothers with families (metros/Tier-1, ₹20L+ household income)

• Cook 8-12+ hours/week, know how to cook but desperately need time back

• Cooking for 3-4 people daily, managing nutrition and varying preferences

• More price-sensitive than couples; view devices as utility, not indulgence

• Want to keep cooking but need help scaling the complexity

(Segments like non-working at-home women, rural India, or households below ₹10L annual income will likely adopt much later - different pain points, different economics.)

The assistance model addresses the larger opportunity because it serves people who want control/taste/experimentation/tech convenience, not just people who don’t want to cook at all. Different value propositions for different segments.

Smart money is entering the category in India

If you observe India’s market map, recent capital deployment patterns signal investor conviction that cooking tech has shifted from an “interesting experiment” to an “investable category.” Here’s what’s happened in the recent 8-12 months:

• June 2025: Upliance AI (AI-powered cooking vessel with guided recipes) raised ₹34 crore / ~$4 million from Khosla Ventures at ₹143 crore / ~$17 million valuation (YourStory, The Arc, 2025).

• August 2025: Beyond Appliances (smart IoT kitchen appliances with digital safety features) raised $4 million / ₹33.5 crore led by Fireside Ventures (Inc42,

  2025).

• August 2025: While not a funding round, this is worth noting as cooking tech companies increasingly target India as a market. Rotimatic entered India with its flagship product priced at ₹1,24,999.

• October 2025: An AI-first cooking companion device (in stealth), raised ~₹3 crore / ~$333k from Antler, and other marquee angels

• 2025: Geek (smart pressure cookers and appliance automation) secured ₹4 crore / ~$480,000 in total funding via Startup Singam (Entrackr, 2025).

Four significant funding rounds in consumer hardware - in one of the hardest categories to fund in Indian venture capital - in a single year isn’t coincidence.

Khosla Ventures’ involvement in this category is interesting. Khosla has a track record of early-stage deep-tech bets that seemed contrarian at the time: Impossible Foods (2011), Square (2009), Affirm (2012). They bet on Instacart in 2012 when grocery delivery seemed impossible to unit economics.

Their thesis typically centres on: (1) technology inflection points that make previously impossible business models viable, (2) large markets with incumbent inertia, (3) founders with deep domain insight.

Fireside Ventures’ thesis is also telling. They focus on consumer brands addressing aspirational consumption in India’s rising middle class. Their portfolio includes Mamaearth, The Whole Truth, Yogabar, Bombay Shaving Company - brands that succeeded by offering premium products to consumers willing to pay for quality and convenience. Their investment in Beyond Appliances signals conviction that kitchen has moved from early-adopter niche to mass-premium category.

Importantly, these companies are carving different niches within an emerging market:

• Upliance: AI-first, full cooking automation, premium positioning (₹29,999 / ~$330)

• Beyond Appliances: Safety-first, IoT-enabled, smart appliances, mass-premium (₹15,000-25,000 / ~$180-300)

• (Unannounced, in stealth): AI-first cooking companion device (₹8,000-12,000 / ~$95-145)

• Geek: Pressure cooker automation, entry-premium (₹8,000-12,000 / ~$95-145)

This product segmentation is itself a signal of category health. But capital alone doesn’t create demand. Let’s dive into tailwinds and micro-needle movers for the buyer today.

Demographics created the buyer

The economic case has become quantifiable for dual-income households. The opportunity cost is measurable. Consider a typical dual-income household:

• Combined household income: ₹15-25 lakh annually / ~$18,000-30,000 (roughly ₹1,250-2,100 per hour combined, assuming 2,000 working hours per year)

• Time spent cooking: 13.5 hours per week

• Weekly opportunity cost: 13.5 hours × ₹1,650 average hourly rate = ₹22,275 per week

• Annual opportunity cost: ₹11.6 lakh / ~$14,000

A ₹40,000 / ~$480 device that cuts cooking time by 40% (saving ~5.4 hours per week) creates ₹8,910 of value per week, or ₹4.6 lakh / ~$5,500 annually. The payback period is roughly 1 month.

The calculation has limitations - time saved doesn’t automatically convert to income earned, and household time allocation is complex. But the directional insight holds: when women’s time has economic value, a ₹30,000 device with one-month payback makes rational sense in a way it never did when that same time was “expected duty” with no price tag.

This shift in economic framing makes a ₹30,000 purchase decision rational. But economics alone doesn’t drive adoption.

There’s a second demographic shift that matters as much: cooking as identity, not just necessity.

“Young Indians who cook don’t cook because they have to - they cook because they want to,” Mahek observes. “This became very clear to us very early. Cooking is an identity for a lot of people. We allow people to do this efficiently.”

And there’s a third shift: palate diversification.

Young India’s palate is changing faster than ever. The traditional dal-roti-sabzi rotation is expanding rapidly. Pasta and noodles on weekdays. Elaborate biryanis on weekends. Korean bibimbap experiments. Healthy high-protein bowls. I don’t even remember the last time I’ve had roti-vegetable curry-dal for 3 days non-stop. This trend is also obvious from the keywords Indians are searching to cook/explore different cuisines:

The influencer economy has amplified this shift. Cooking content in India is exploding:

• Hebbar’s Kitchen: 8.03M YouTube subscribers, clocking 15M views in the last 30 days alone [Social Blade, account analytics tools]

• Cookd: 3.9M YouTube subscribers, closing 79M views in the last 30 days alone [Social Blade, account analytics tools]

• Kabita’s Kitchen: 14.6M subscribers, specialises in traditional Indian recipes

• Cook With Nisha Madhulika: 15.1M subscribers on YT, one of India’s OG veg food creators [Social Blade, account analytics tools]

• Your Food Lab: 7.02M subscribers on YT, fusion and innovative cooking [Social Blade, account analytics tools]

• Rajshri Food: 3.36M subscribers on YT, hit nearly 663M views across all videos, traditional recipes with modern twists [Social Blade, account analytics tools]

These creators make diverse cooking look accessible. They break down intimidating cuisines into achievable steps. To my surprise (not really), I have 300 cooking videos saved in a folder on Instagram, and I only cook 4-5 hours a week. And I’m no longer eating chapatis, traditional vegetable curries, and dal daily. My week is a mix of chilli chicken, tomato basil pasta, salad, air fried calzones, and what not.

The gap between “what I want to eat” and “what I know how to cook” has widened dramatically. Food delivery bridged this gap temporarily - but delivery is expensive (₹300-500 per meal including delivery fees and tips) and often unhealthy. The desire exists to make these dishes at home. Time and skill don’t.

Korean bibimbap from scratch takes 90 minutes of prep. Thai green curry requires ingredients most households don’t stock and techniques they haven’t practised. The aspiration exists. The activation energy is too high. This creates the white space for cooking tech: enable variety without the learning curve. Make the Instagram dish achievable on a Tuesday night.

The behavioural shift happened before the technology arrived. Young Indians wanted to cook differently. The technology is catching up to existing demand.

Critically, these three factors - technology maturation, capital availability, and demographic shift - aren’t independent. They compound:

The technology maturation makes the category investable → Capital enables manufacturing scale and category marketing → Scale reduces prices and increases distribution → Demographic demand drives adoption → Usage data improves AI models → Technology improves further.

This is the flywheel that previous attempts lacked. Each element reinforces the others.

III. The next 5-10 years

If the technology delivers on reliability and adoption follows predicted curves, Indian kitchens could transform over the next decade. But “could” requires careful qualification. Here’s what the trajectory might look like - with attention to what could go wrong.

The time reallocation (If adoption happens)

My prediction is that by 2030, households with AI/smart cooking tech could spend 8-9 hours per week cooking instead of 13.5. That’s a 33-40% reduction. Still above the global 6-hour average, but meaningfully lower.

This estimate draws from Indian appliance adoption patterns:

• Pressure cookers (1950s-1980s): Transformed Indian kitchens by reducing dal and rice cooking time by 60-70% - what took 45-60 minutes now took 15-20 minutes (Kitchendemy, 2024). Achieved near-universal urban adoption within 30 years of mass availability. Introduced commercially in India in 1925, reached scale post-1950s with brands like Hawkins and Prestige (The Better India, 2020; Medium, 2022). Why? Immediate daily utility, affordable pricing (₹500-1,500), and minimal infrastructure needs.

• Washing machines (1950s-2025): Despite launching in the 1950s-60s, penetration remained just 11% by 2016 and is projected to reach only 21.8% by 2026 (Statista/Edelweiss Research, 2017). Why so slow? High prices (₹10,000-15,000+ for decades), infrastructure requirements (water supply, drainage, electricity), and competition from domestic help.

Cooking tech sits between these two analogies. Like pressure cookers, it addresses a high-frequency daily pain point with immediate time savings. Like washing machines, it requires infrastructure (counter space, electricity) and competes with low-cost alternatives (domestic help, food delivery). Current pricing (₹25,000-40,000) suggests slower adoption than pressure cookers unless prices drop to ₹15,000-20,000 by 2027-28.

The tedious components get automated - chopping vegetables, constant stirring, timing management across multiple dishes, temperature monitoring. The creative parts stay human - deciding what to cook, seasoning adjustments, plating, family preferences.

But here’s the counterintuitive part: some saved time might flow back into cooking. Not as drudgery, but as experimentation. More elaborate weekend cooking when you have time to enjoy the process. More festival cooking for celebrations. The ratio of “joyful cooking” to “obligatory cooking” shifts.

The kitchen transforms from “have to be there” space to “want to be there” space - but this transformation will take 10-15 years, not 3-5.

But this assumes sustained adoption and continued product improvement. What does the market trajectory look like?

Market growth at two speeds

To have a glimpse of the cooking tech market map of India, and other markets such as China, Europe & the US, click here.

Cooking tech/smart appliance in India will grow at 20-30% annually for the next 2-3 years, then moderate to 15-20% as the category matures. This is faster than the overall category because they’re starting from a small base (classic CAGR compression - going from 100 units to 200 units is 100% growth, but going from 10,000 to 15,000 is only 50% growth even though you added more units).

For context on comparable categories in India:

• Smart speakers have shown explosive growth, with current projections ranging from 17-34.28% CAGR depending on the forecast period and source (IMARC, Expert Market Research 2024-2025).

• Air fryers have established steady growth now at approximately 5.8-5.9% CAGR (TechSci Research, BlueWeave, IMARC 2024-2025).

• Smart home devices have also shown strong adoption curves with a CAGR of 31.6% through 2030. A standout example of this rapid uptake is the Smart Security & Safety segment, which commanded a massive 45.1% share of all smart home demand in 2025, as Indian consumers pivot toward “safety-first” automation (Source: Grand View Research / Mordor Intelligence, 2026).

Cooking tech likely follows a similar trajectory in its early years (high growth as category establishes) before moderating to steady growth rates.

Pricing trajectory and market expansion

The current price point is ₹29,999 (~$330) for Upliance (although it’s on sale at ₹25,999 / $289). That’s mass-premium positioning - aspirational but accessible for dual-income urban households in metros and tier-1 cities.

Price trajectory for cooking tech companies will depend heavily on:

• Manufacturing scale: Consumer electronics typically see 15-20% cost reduction for every doubling of cumulative production volume (Wright’s Law / experience curve). As volumes increase from thousands to tens of thousands, unit costs decline.

• Competition intensity: Currently limited competition at the AI-first tier. If traditional appliance makers (Philips, Morphy Richards, Prestige) enter aggressively, or if Chinese manufacturers (like Xiaomi, who’ve succeeded in smart home devices) enter India’s market, pricing pressure intensifies.

• Component costs: AI chips, sensors, and motors are the cost drivers. These are declining every year due to semiconductor improvements and sensor commoditisation.

Realistic price trajectory:

• 2025-2026: ₹35,000-40,000 (~$420-480) - current positioning holds

• 2027-2028: ₹25,000-30,000 (~$300-360) - competition drives prices down, scale improves economics

• 2029-2030: ₹18,000-22,000 (~$215-265) - mass-market positioning becomes viable

By 2027-2028, at ₹25,000-30,000, the product becomes accessible to upper-middle-class households in tier-2 cities (Jaipur, Lucknow, Indore, Coimbatore). This is the inflection point for market expansion.

By 2030, at ₹10,000, tier-3 penetration begins, though slowly. At these prices, the device competes with established kitchen appliances (mid-range mixer-grinder + pressure cooker + air fryer ≈ ₹15,000 combined). The value proposition becomes “replace 3 devices with 1 and get AI assistance.”

But price isn’t the only barrier. Trust, category education, and behaviour change matter as much.

Category formation signals (what to watch)

The next three years determine whether this becomes a real category or a niche product for affluent early adopters. Here are the signals that distinguish category formation from category failure:

Signal 1: Kitchen appliance e-commerce growth rate divergence

Watch for online channels growing faster than offline channels for kitchen appliances.

Digital channels for kitchen appliances grew 25% in 2024, significantly outpacing overall market growth of 6-7% (Mordor Intelligence citing India Infoline, 2025). Online retail sales of kitchen appliances reached ₹3,750 crore / ~$450 million in 2023 (Ken Research, 2025). However, e-commerce still represents a relatively small portion (9-30%) of total kitchen appliance sales, indicating substantial room for expansion.

What to track: Year-over-year growth rates for online vs offline kitchen appliance sales. E-commerce penetration as % of total kitchen appliance sales.

Signal 2: EMI adoption and price-point segmentation

22% of Upliance sales come via EMI, positioning it between appliances and aspirational electronics. This indicates buyers view it as a considered purchase - more than discretionary, not yet essential.

My survey reveals three price-sensitivity tiers that map directly to distinct buyer personas (age - 28-35):

• Premium tier (₹35K+): time-starved couples

  • 7% of respondents willing to pay ₹35K+

  • Profile: Young couples, dual-income, ₹2L-4L+ household income

  • Primary motivation: “Reduces mental load (planning, guidance)” - 44% cite this

  • EMI willingness: 21% of couples willing to finance at this tier

• Mid-tier (₹20-35K): skill-gap aspirants

  • 20% of respondents in this range

  • Profile: Living alone or young families, ₹1L-2L income, cooking 0-3 hours/week but wanting more

  • Primary motivation: “More consistent/better results” and “helps with variety”

  • EMI willingness: 17% willing to finance

• Mass-market tier (Under ₹20K): efficiency maximisers

  • 73% of respondents

  • Profile: Families with kids, cooking 8-12+ hours/week, ₹50K-2L income

  • Primary motivation: “Saves significant cooking time” - 41% cite this

  • EMI willingness: Only 12% willing to pay ₹20K+ (highest price sensitivity)

Upliance at ₹29,999 targets Premium Tier couples paying for mental load reduction. The 22% EMI adoption shows Mid-Tier aspirants stretching into this segment through financing.

This creates sequential unlocking:

• Years 1-3 (current): Capture Premium Tier at ₹35-45K with “convenience + comprehensive solution” messaging

• Years 3-5: Price compression to ₹20-25K opens Mid-Tier with “confidence-building + variety” messaging

• Years 5-8: Further compression to ₹10-15K reaches Mass-Market with “time efficiency” messaging

What to track:

• EMI adoption by household type: Are couples financing at 20%+ whilst families stay below 15%? Validates Premium Tier exists and Mid-Tier is stretching upward.

• Income bracket of EMI users: If ₹1L-2L buyers use EMI at 30%+ whilst ₹4L+ pay cash, middle-income is financing to access premium tier - strong demand signal.

• Purchase motivation shifts: As prices drop, does “mental load reduction” give way to “time-saving”? Shows successful transition from Premium to Mass-Market positioning.

• Competitors’ price-tier entry: New players at ₹15-25K prove Mid-Tier exists. Watch if they position on confidence-building vs efficiency.

Signal 3: Tier-2/3 early adoption

Are smart cooking devices appearing in Jaipur, Kochi, Indore searches and purchases before prices drop significantly?

Upliance already sees 25% of sales from tier-2 cities. More tellingly, “Tier-2 actually has better engagement metrics,” Mahek notes. “They watch more of the recipe videos, they use it more frequently, they’re less likely to return it.” Her perspective: “I like the tier-2 market. People want to do more things themselves. You’ll be surprised by the ability to spend if people were to make products that work for them.”

This is a strong positive signal - there’s a growing customer pool in tier-2 isn’t waiting for lower prices, and they’re more engaged than metro users.

What to track: Geographic distribution of sales - metro vs tier-2 vs tier-3. Early tier-2 adoption (even at 15-20% of sales) before pricing drops indicates category pull, not just metro trend.

Signal 4: Gifting category emergence

Are cooking tech devices showing up as premium gifts for Diwali, housewarmings, weddings?

Diwali 2024 saw some cooking tech gifting, but volumes were small. This is a watch area for Diwali 2026-2027.

In Indian consumer behaviour, gifting category signals aspirational positioning. Gold, electronics, premium appliances become gifts when they signal status and thoughtfulness. If cooking tech enters this category, it’s transcended “functional appliance” to become “lifestyle product.”

What to track: Seasonal sales spikes around Diwali/Dhanteras (October-November) and wedding season (November-February). If cooking tech shows 40%+ spike during festival season (similar to electronics), it’s entering gifting category. If spikes are <20%, it’s still functional-only purchase.

Signal 5: Organic content creator adoption

Are food YouTubers and Instagram creators showcasing these devices? Not sponsored posts - organic usage.

Limited organic adoption so far. Most creator content is sponsored reviews. This is actually a yellow flag - if the product were transformatively useful, creators would adopt it organically.

What to track: Top 20 Indian food creators’ content. Are they using AI cooking devices in regular recipe videos (not just sponsored reviews)? What’s engagement on those videos versus their average?

Signal 6: Marketing language evolution

Watch for messaging shift from “time-saving for busy professionals” (burden framing) toward “explore new cuisines” and “cook what you love” (aspiration framing).

Mixed messaging. Some companies emphasise “joy of cooking” and “explore cuisines” (good), but some communication still leans on “save time” (functional). This needs to shift further toward aspiration.

Products framed as solving a burden may sell to early adopters but struggle to achieve mainstream cultural acceptance. Products framed as enabling aspiration scale faster. Compare: washing machines initially marketed as “reducing burden” (practical but unexciting) versus being re-framed as “freeing time for what matters” (aspirational).

What to track: Brand campaigns, website copy, ads from Upliance, Posha, Nosh, and new entrants.

Signal 7: Traditional appliance makers’ response

When do incumbents (Philips, Prestige, Morphy Richards, Bajaj, Butterfly) launch competitive products?

Wonderchef launched “Chef Magic” in 2024, but it’s more IoT-enabled than AI-first. Traditional players are watching, not aggressively competing yet. This could change in 2026-2027.

Incumbents moving fast signals they see category threat. Incumbents waiting signals they’re skeptical of market size. Fast follower entry creates competition (bad for margins) but validates category (good for overall growth).

What to track: Product announcements from traditional players. Do they launch AI-powered cooking devices in 2025-2026, or wait 2-3 more years?

IV. For investors and founders

Three important patterns matter for understanding whether cooking tech succeeds in India versus following previous failed attempts.

Product-market fit requires cultural fit

Indianisation isn’t optional. It’s the entire game.

Thermomix succeeded in Europe and Australia with Western recipes. It’s been available in India for years (₹2-4 lakh / ~$2,400-4,800 price range). It will never achieve mass adoption here because:

• Recipe library optimised for European cooking (bolognese, quiche, breadmaking)

• Assumed ingredients available in Western supermarkets (specific cheeses, pre-made stocks)

• Didn’t handle Indian cooking’s judgment-based techniques (tadka, dum, bhapa)

Upliance’s initial traction comes from approach but for Indians: 750+ Indian recipes from the start. More importantly, it allows for Indian cooking’s variability: “Add ghee instead of oil,” “Make it less spicy,” “I don’t have tomatoes, can I substitute?”. The AI handles these adjustments rather than failing when users deviate from the recipe.

For founders

If you’re building in this space, the moat isn’t technology alone. It’s cultural and culinary understanding embedded in the product. You’re competing with decades of cooking knowledge that exists in Indian homes. Your product needs to respect that knowledge, not replace it.

For investors

Evaluate whether teams have genuine insight into Indian cooking behaviour. The CEO or founding team should either cook regularly themselves or have deeply embedded observation of how their target users cook. Products built by people who view Indian cooking as “exotic” or “complicated” will miss nuances that matter.

Joy positioning beats efficiency positioning

This is counterintuitive but critical.

Rationally, you’d think efficiency messaging works best for time-starved dual-income households: “Save 5 hours per week. Get your time back.”

But consumer behaviour research consistently shows: people don’t buy efficiency for its own sake. They buy aspiration and identity.

Compare:

• Peloton doesn’t sell “efficient exercise” (though it is). It sells “become who you want to be.”

• Apple doesn’t sell “efficient smartphones” (though they are). It sells “tools for creative people.”

• Dyson doesn’t sell “efficient vacuuming” (though it is). It sells “beautiful engineering for your beautiful home.”

Cooking tech needs similar framing. Not “save time on cooking,” but:

• “Cook like you travelled to that place”

• “Make restaurant-quality dishes at home”

• “Explore cuisines you’ve been intimidated by”

• “Become the cook you always wanted to be”

My survey data also reveals why this framing works. When asked about the most frustrating cooking moments, responses clustered around anxiety and failure: burning food after forgetting the gas, output not matching recipe, constantly monitoring to prevent mistakes. When asked about most enjoyed moments, responses clustered around mastery and satisfaction: “the whole process feels like therapy,” “the after glow of a perfectly done dish,” “when I taste it and it’s exactly what I wanted.”

The divide isn’t tedious vs enjoyable cooking. It’s anxiety-inducing unpredictability versus satisfying mastery. People hate the fear of wasting time and ingredients on something that might fail. They love the confidence of creating something impressive.

The efficiency is real and valuable - these devices do save 40% of cooking time. But efficiency is the mechanism, not the message. The device removes anxiety-inducing failure points (timing, temperature, monitoring) so users can focus on the identity-affirming parts (experimenting, creating, impressing others).

This explains Mahek’s observation about tier-2 cities showing higher engagement. These users view successfully cooking diverse restaurant-style dishes as an aspirational skill. The device doesn’t replace cooking - it makes them confident enough to attempt dishes they’d otherwise order out.

Mahek puts it directly: “Price needs to come down, but that is not the primary battle. First we need to make cooking cool and diverse.”

For founders

Audit your marketing. If it leads with time-saving and burden reduction, you’re solving a problem, but not creating desire. Reframe around aspiration: what does your product let people become?

For investors

Watch how companies talk about their product. Are they selling against a burden (cooking is hard, time-consuming, stressful) or toward an aspiration (explore, experiment, delight)? The latter has much higher willingness-to-pay and category expansion potential.

The category formation window Is short

Three years. Maybe four. That’s the window for category leadership to be decided.

Smart speaker market in India consolidated quickly once Amazon Echo and Google Home established positions (2017-2020). Today, new entrants struggle for mindshare even with better products because “smart speaker = Alexa or Google Home” is locked in consumer consciousness.

Cooking tech likely follows similar dynamics. Early category leaders benefit from:

• Brand = category association (like “Xerox” or “Google it”)

• Data advantages (more users → more cooking data → better algorithms → better product)

• Manufacturing scale (early volume → better unit economics → ability to price aggressively)

• Distribution partnerships (first mover gets better shelf space, better retailer terms)

The window is now because:

• Technology maturation: AI/tech reliability is sufficient for mass-market launch

• Demographic timing: dual-income households growing fastest in next 3-5 years

• Capital availability: funding environment favours consumer tech with clear use cases

• Competitive landscape: traditional appliance makers haven’t moved aggressively yet

In 3-4 years, if category forms successfully:

• Incumbents will launch competitive products with distribution advantages

• Chinese manufacturers may enter with aggressive pricing

• Market will have clearer winners, making new entry much harder

For founders

If you’re building in this space, now is the time. In 2027-2028, you’ll be competing with established players with 2-3 years of user data and manufacturing scale. Today, you’re all starting from similar positions.

For investors

The next 18-24 months are the deployment window. If you wait for “proof” of category formation, you’ll miss the category leaders. Early bets on teams with genuine insight carry more risk but vastly more upside.

V. What could go wrong (risks that matter)

As an investor analysing this space, I see five categories of risk that could prevent category formation or significantly delay the timeline I’ve outlined. These aren’t reasons to avoid the space - they’re factors to track as signals of whether this thesis is playing out as expected.

Price elasticity might be much lower than expected

The fundamental question: will mainstream consumers wait for ₹15,000-20,000 devices, or will they buy at ₹35,000-40,000?

Current data is limited to early adopters. These buyers have high willingness-to-pay and are less price-sensitive. As the product moves from early adopters (innovators + early majority) to mainstream majority, price sensitivity typically increases sharply.

EMI financing might bridge this gap - but only if consumers view the product as “essential enough to take a loan for.” That’s unproven.

If mass-market adoption waits until 2028-2029 (when prices reach sub ₹15,000), the addressable market remains small for next 3-4 years. That could:

• Limit manufacturing scale, keeping costs high

• Reduce investor enthusiasm for follow-on rounds

• Delay the flywheel of scale → lower prices → more adoption → better AI → better product

Trust builds slowly, breaks instantly

One high-profile product failure - a device that starts a fire, burns food repeatedly, causes injury - could set the category back 18-24 months. (Reminds of EV scooters when they entered the market, that’s exactly the risk I’m talking about)

The kitchen’s high stakes cut both ways. Users demand reliability. But early-stage products often have quality issues. In consumer electronics, this is typically non-catastrophic. In kitchen appliances, more’s at stake beecause quality issues are:

• More visible (burnt food, bad taste, wasted ingredients)

• More emotionally impactful (cooking for family, guests, special occasions)

• Potentially dangerous (fire risk, burns, electrical issues)

Behaviour change is hard (the peloton problem)

Perfect technology doesn’t guarantee daily adoption. If I have the perfect cook or pre-cooked rajma at my disposal, my chopping has become easy, will I adopt anything else?

Pelotons gather dust in bedrooms. Treadmills become clothes racks. Home exercise equipment has notoriously poor sustained usage rates - 70-80% of buyers stop using equipment within 6 months of purchase.

Cooking devices could face similar dynamics:

• Initial enthusiasm: buyer is excited, tries 5-8 recipes in first month

• Novelty wears off: back to familiar cooking methods, device sits unused

• Guilt and rationalisation: “I should use it more” but never do

• Eventual abandonment: device takes up counter space, gets stored away

Chaitanya acknowledges this challenge candidly: “There is a challenge of long-term retention - like most devices, how do you keep them coming back after the initial honeymoon period? And even if they enjoy the product a lot, sometimes they can ‘graduate’ to cooking on the stove. Or hire a cook later for some reason and never get back.”

The 3.42 meals per week average represents active users - but the key unknown is what percentage of buyers remain active after 6, 12, 18 months. This is the metric that determines whether the device becomes a kitchen staple or joins the air fryer in the back of the cabinet.

For this risk to manifest: 30-day retention is high (80%+), but 6-month retention drops to 30-40%. Users who cook 3-4 times per week initially drop to less than once per week.

Incumbents move fast with distribution advantages

Traditional appliance makers (Prestige, Philips, Bajaj, Morphy Richards) have:

• Established distribution (present in 50,000+ retail stores)

• Brand trust (decades of reputation)

• Manufacturing scale (existing production facilities)

• Customer relationships (existing user base to cross-sell)

If they launch competitive AI cooking products in 2025-2026, they can leverage these advantages to:

• Price aggressively (they can sustain lower margins due to scale)

• Achieve instant distribution (no need to build retail presence from scratch)

• Bundle offerings (buy pressure cooker + mixer + smart cooking device at discount)

• Market to existing customers (cheaper customer acquisition than startups)

For this risk to manifest: By 2026, Prestige or Philips launches AI cooking device at ₹28,000-32,000 with comparable features. They leverage retail presence to capture 30-40% market share within 12 months.

Unit economics may be difficult at scale

Hardware is brutally difficult. Unit economics that work at 5,000 units/year may break at 50,000 units/year when:

• Return rates increase with faster production (quality control issues)

• Support costs scale faster than anticipated (more users = more issues)

• Marketing costs remain high (customer acquisition doesn’t get cheaper as fast as hoped)

• Distribution costs increase (offline retail has different economics than online DTC)

For this risk to manifest: Companies raise Series A/B on promising early traction, then struggle to achieve profitability even at scale because LTV:CAC ratios are worse than projected.

Why this time it’s different

You read how India’s palate is changing. How young couples are Googling “Korean fried chicken at home” after ordering it on Swiggy. How working mothers/parents are time starved to cook around the clock. The aspiration today has shifted from mastering what your mom made to experimenting with what you saw on Instagram - and that gap between wanting to cook something and confidently executing it? That’s where this opportunity lives.

Every big brand that cracked Indian kitchens started with convenience as the hook. Maggi didn’t teach you to cook better - it gave you 2 minutes. Instant Pot in the west didn’t reduce judgment calls - it assumed you already knew what you were doing. Microwave ovens couldn’t handle tadka. The promise was always “save time,” never “cook what intimidates you with ease/joy.” Nobody solved for the actual ‘physical’ and ‘mental’ load cooking brings with it. Until now.

The tech finally works, and it’s happening on two fronts.

Infrastructure: For the first time, companies like Cibos are building DC heating elements and thermal components in India. We don’t need to retrofit Chinese imports. That means appliances can actually execute on how Indians cook.

AI: Vision-language models can now watch oil separate in a kadhai and adjust heat in real-time. Multimodal AI fuses visuals, temperature, weight, and sound to make judgment calls - ”this dal needs two more minutes,” “these onions need higher heat.” Edge computing chips dropped to consumer pricing, so this runs on-device without lag.

The TAM is real, and it’s big.

Based on the 23-24 ITR statistics, 6.6 million (0.66 Crore) Indians earn above ₹15 lakhs annually. Within that:

• 3.5 - 4M dual-income nuclear families

• 1.5 - 2M DINK households

• Total addressable: ~4.5 - 5M households (Premium)

At ₹25-30K pricing = ₹11,250 - 15,000 crore (~$1.3 - 1.8B) right now. 13 million (1.30 Crore) Indians earn above ₹10 lakhs annually.

• Total addressable: ~9 - 10M households (Mass Premium)

• The Shift: As pricing drops, this captures 2x more households as Tier-2 accelerates.

• At ₹15-20K pricing (2028-2030) = ₹13,500 - 20,000 crore (~$1.6 - 2.4B).

This isn’t “everyone who cooks.” It’s dual-income households where cooking time has measurable economic value. That base is growing at 22% CAGR as dual-income penetration rises from 42% to 55%+.

Two emerging ICPs are already buying:

Young couples and DINKs (₹20L+ dual income, 4-7 hrs/week cooking) want mental load reduction and experimenting with cuisines they’ve been intimidated by. 21% willing to pay ₹20K+. They’re the premium segment buying to cook the Pav Bhaji they saw on Reels without anxiety.

Working mothers with families (₹20L+ household, 8-12+ hrs/week cooking) need sub-₹15K pricing. But here’s the tell: Tier-2 cities already show 25% of Upliance’s sales with better engagement than metros. When pricing drops, this segment explodes.

And it’s working. Upliance users are averaging 3.42 meals per week. That’s habit formation in the making. 22% of buyers are opting for EMI financing on a ₹30K device, treating it like infrastructure, not an impulse gadget. Top recipes aren’t basic dal-chawal - they’re restaurant-style Pulav and Pav Bhaji. People are cooking things they previously outsourced or avoided.

Four major funding rounds happened in 2024-2025 alone: Upliance (₹34 crore from Khosla), Beyond Appliances ($4M from Fireside), stealth cooking AI (₹3 crore from Antler), Geek (₹4 crore). In consumer hardware. In India. In one year.

The window to build, innovate, and invest in this space is now. By 2027-2028, Prestige and Philips leverage distribution, Chinese manufacturers bring aggressive pricing, and early movers have the data moat and manufacturing scale that late entrants can’t buy.

Our kitchens are about to get its Alexa moment.

The generation that grew up watching their mothers spend entire Sundays in the kitchen will spend their Sundays differently. They’ll still cook. They’ll just cook what they want to cook, not what they have to cook, with joy.

In 5 years, the 13.5-hour gap will come down by half. Not because Indians will stop cooking. Because cooking will stop being tedious while remaining authentic. Because technology finally gave an entire generation permission to cook on their own terms.

The ₹20,000 crore question isn’t whether this happens.

It’s who builds it - and whether they understand this isn’t about replacing cooking. It’s about making mastery accessible. And maybe, just maybe, part 2 of this blog will be about walking into my kitchen without that familiar knot of frustration. About being a 26-year-old woman who finally has the tech that centuries of women deserved but never got.

P.S. If you're taking on hard, ignored problems like cooking tech - in consumer AI, or anywhere else with AI - happy to jam! I’m at dravisha.katoch@antler.co.

Acknowledgements

This wouldn’t exist without some people. A big thank you to:

Kuldeep (Last9), Sapna (WaterBridge), Gowri (Antler), and Rajat (my husband) for reading my messy drafts. Pranav (Accio Robotics) for the introductions. Mahek (CEO, Upliance AI) and Chaitanya (ex-Head of Product, Upliance AI) for being candid with data, and thoughts. Prithvi (ex-Posha) and Aditya (Cibos) for sharing insights on tech and infra.And to the 58 people who filled out my primary survey - thank you.

Sources and References

Cooking Time Statistics

• GfK Global Study. “Time Spent on Cooking: A Global Comparison.” GfK Consumer Life Study, 2019.

• Open City. “Childcare and Domestic Work - Time Use Survey: Gender Disparities in Urban India.” Open City Foundation, 2020.

Historical Market Data

• Census of India. “Household Composition and Economic Activity Data.” Census 2011, Government of India.

• Morgan Stanley. “The Dual Income Household in India: Rural Transformation and Urban Growth.” Morgan Stanley Research, 2024.

• Income Tax Department. “Annual Report 2023-24.” Ministry of Finance, Government of India, 2024.

• Reserve Bank of India. “Urban Income Distribution Study 2023.” RBI Research, 2023.

• CRISIL. “Consumer Study 2024: Household Structure and Spending Patterns.” CRISIL Research, 2024.

Financial Inclusion & Women’s Economic Participation

• World Bank. “Global Findex Database 2021: Financial Inclusion, Digital Payments, and Resilience in the Age of COVID-19.” World Bank Group, 2021.

• Ministry of Statistics and Programme Implementation (MoSPI). “Women and Men in India 2024.” MoSPI, Government of India, 2024.

• Bain & Company India. “Female Labour Force Participation in India: Trends and Drivers.” Bain India Reports, 2024.

• TransUnion CIBIL. “Credit Growth and Women Borrowers in India.” TransUnion CIBIL Credit Market Indicators, 2024.

• NITI Aayog. “Women Entrepreneurship and Financial Inclusion.” NITI Aayog Reports, Government of India, 2024.

• SBI Research. “Women Investors in Indian Equity Markets: A Demographic Shift.” State Bank of India Economic Research Department, 2024.

• DBS Bank and CRISIL. “Financial Decision-Making in Urban Indian Households: The Role of Women.” DBS-CRISIL Joint Survey, 2024.

Appliance Adoption & Kitchen Technology History

• Statista. “Penetration rate of the washing machine market in India from 2005 to 2016, with estimate till 2026.” Edelweiss Research, July 17, 2017.

• The Better India. “Icmic, Santosh & Rukmani: The Forgotten Story of India’s Original Cookers.” The Better India, July 11, 2020.

• Sarkar, Nabamita. “The Age Old Tale Of Indian Pressure Cookers.” Medium, September 4, 2022.

Consumer Behaviour & Purchasing Patterns

• Fortune India. “The Rise of Female Consumer Power in India.” Fortune India, TechCrunch coverage, 2023.

• Home Credit India. “How India Borrows: Consumer Credit and Shopping Behaviour Report 2024.” Home Credit India, 2024.

• IPG Mediabrands. “Gender Parity in E-commerce: India’s Digital Shopping Revolution.” IPG Mediabrands Research, Brand Equity, 2024.

• Indian Retailer. “E-commerce Growth Trends: The Female Shopper Surge.” Indian Retailer Industry Report, 2024.

• Flipkart. “Flipkart Trends 2025: Consumer Insights and Shopping Behaviour.” Flipkart Commerce Insights, 2025.

Food Industry & Consumer Trends

• Nestlé India. Annual Reports and Investor Presentations, 2010-2024.

• Finception. “Maggi: Nestlé India’s Crown Jewel - A Financial Analysis.” Finception Industry Research, 2018.

• OutlookBusiness. “Nestlé India’s Product Portfolio Performance 2025.” OutlookBusiness.com, 2025.

• buildd.co. “Indian FMCG Brands Market Share Analysis.” Buildd India Market Intelligence, 2024.

• Business Standard. “Nestlé India Revenue and Brand Performance Q4 2024.” Business Standard, 2024.

• Eastlake (via FoodNavigator). “Global Home Cooking Trends: Health, Convenience, and Consumer Preferences.” Eastlake Consumer Research reported in FoodNavigator, 2023.

Startup Funding & Industry Coverage

• YourStory. “Upliance AI Raises $4 Million from Khosla Ventures.” YourStory Startup Coverage, June 2025.

• The Arc (formerly The Ken). “Khosla’s Bet on Indian Kitchen Tech: Upliance AI Funding Analysis.” The Arc, 2025.

• Inc42. “Beyond Appliances Secures $4 Million in Series A Funding from Fireside Ventures.” Inc42 Startup News, August 2025.

• Entrackr. “Geek Raises ₹4 Crore to Scale Smart Kitchen Appliances.” Entrackr Funding Coverage, 2025.

Technology & Market Analysis

• Mordor Intelligence. “India Kitchen Appliances Market: Growth, Trends, and Forecasts (2024-2033).” Mordor Intelligence Industry Reports, 2024.

• Euromonitor International. “Consumer Appliances in India: Country Report.” Euromonitor Passport Database, 2024.

Primary Research

• Exclusive interviews with:

  • Mahek, Co-founder & CEO, Upliance AI (November 2024)

  • Chaitanya, former Head of Product, Upliance AI (November 2024)

• Primary survey: “Cooking Habits and Smart Kitchen Tech Adoption.” 58 respondents, convenience sample via author’s network, November-December 2024. Survey distributed via Google Forms.