BREAKING: Tanay Jaipuria tracks the 240x collapse in AI inference costs - and turns it into must-read weekly analysis ★ PARTNER AT WING VENTURE CAPITAL ★ FORMER META PM ★ 72K+ FOLLOWERS ON X ★ TANAY'S NEWSLETTER: The essential Substack for anyone who wants to understand AI economics before the market does ★ BAKER SCHOLAR, HARVARD MBA ★ CS DEGREE, COLUMBIA ★ MUMBAI TO NEW YORK ★ BREAKING: Tanay Jaipuria tracks the 240x collapse in AI inference costs - and turns it into must-read weekly analysis ★ PARTNER AT WING VENTURE CAPITAL ★ FORMER META PM ★ 72K+ FOLLOWERS ON X ★ TANAY'S NEWSLETTER: The essential Substack for anyone who wants to understand AI economics before the market does ★ BAKER SCHOLAR, HARVARD MBA ★ CS DEGREE, COLUMBIA ★ MUMBAI TO NEW YORK ★
Tanay Jaipuria - Partner at Wing Venture Capital
Profile / Tech & Product

Tanay
Jaipuria

Partner, Wing Venture Capital

The analyst who tracked AI's cost collapse before anyone else was measuring it - and built a 12,000-subscriber newsletter to prove it.

72K+ X Followers
12K+ Newsletter Subs
240x AI Cost Drop He Tracked
15+ Startups Backed
Cover Story

The Analyst Who Won't Let You Sleep On AI Economics

There is a very particular kind of person who looks at a technology market in chaos - in this case, generative AI - and decides the right response is clarity. Not hot takes. Not doom threads. Clarity. That's Tanay Jaipuria: Partner at Wing Venture Capital, former Meta product lead, and the person most likely to have already published a rigorous framework on whatever AI news broke this morning by the time you've finished your coffee.

His newsletter, Tanay's Newsletter, went from side project to essential reading without a single viral moment you can point to. It grew because the analysis was actually useful. Week after week, he turned complex dynamics - the plummeting cost of AI inference, the economics of background agents, the anatomy of defensible moats in an LLM world - into frameworks that operators, investors, and founders genuinely rely on. Twelve thousand subscribers don't show up for vibes. They show up for the math.

The origin story of Tanay Jaipuria is something like a carefully constructed case study in intellectual arbitrage. He grew up in Mumbai, earned a computer science degree at Columbia, spent time at McKinsey advising across financial services and media, then did an MBA at Harvard Business School - where he graduated as a Baker Scholar, placing him in the top 5% of his class. That rare combination of technical grounding, strategic consulting, and elite business education was the foundation. Then Meta handed him a product manager role on News Feed ranking and ad products - the kind of role where you learn, at scale, exactly what makes people pay attention and what makes them click away.

"My guess is that most of the real leverage from AI will come from those quiet loops: agents working in the background while we are in meetings or asleep, and then appearing, briefly, with something that is actually useful." - Tanay Jaipuria

Wing Venture Capital, where he became Partner in 2022, focuses on AI-powered applications, data infrastructure, verticalized SaaS, and product-led growth. Tanay's investment lens is shaped by exactly the operating experience that makes most VC partners credible when they say they "add value beyond the check." He has actually built recommendation systems. He has actually argued for product prioritization inside a company with billions of users. When he tells a founder that their retention curve looks concerning, it's because he's been in rooms where those curves determined whether a product survived its next review cycle.

The newsletter is where you see the full shape of his thinking. He writes about Studio Ghibli film recommendations in the same publication where he breaks down ServiceTitan's S-1. He'll pivot from deep analysis of Reddit's data licensing strategy to a framework on AI browser economics without missing a beat. There's no topic he treats as beneath the rigor he'd apply to a board deck - which is either a symptom of genuine intellectual range or a very disciplined content strategy. Probably both.

His most cited work is the kind of thing analysts usually only discuss in private. When he published his tracking of GPT-4 equivalent intelligence costs dropping 240x in 18 months, it became one of the most referenced data points in AI economics. Not because it was surprising - the trend was visible - but because he was the one who actually measured it and put a number on it. That's the Jaipuria move: take a thing everyone vaguely senses, apply proper analysis, and make it undeniable.

He now operates across New York, thinking carefully about what makes software businesses defensible when AI compresses the labor advantages that used to justify moats. His answer is characteristically precise: AI can't fabricate marketplace liquidity, courier density, reputation history, or a canonical identity graph. Density and trust are structural, not labor-based. That's the kind of insight that sounds obvious in retrospect and non-obvious before someone says it clearly - which is, of course, the entire value proposition of Tanay Jaipuria.

Career Arc

From Mumbai to Wing VC: A Timeline

2008
Joined Twitter/X - making him an early adopter of the platform where he'd eventually build 72K+ followers
2010-2014
Studied Computer Science at Columbia University - the technical foundation everything else gets built on
2014
Joined McKinsey & Company as a consultant, advising clients across financial services, media, and technology
2018
Started MBA at Harvard Business School - would graduate as Baker Scholar (top 5%)
2020
Joined Meta as Product Manager - led teams on News Feed ranking, ad experiences on Facebook, and creator monetization on Instagram
2022
Became Partner at Wing Venture Capital. Launched Tanay's Newsletter on Substack simultaneously
2023
Published "The Plummeting Cost of Intelligence" - tracked 240x drop in GPT-4 equivalent costs over 18 months, became widely cited
2024
Newsletter surpassed 12,000 subscribers. Appeared on Company Breakdowns podcast with Erik Torenberg discussing Reddit's IPO
2025
Published "A Few Themes for 2026" and continued high-impact analysis on AI agents, moats, and enterprise software evolution
Analysis

Six Things That Define the Jaipuria Playbook

01
He Measures What Others Estimate
When the AI cost debate was all vibes and speculation, he tracked the actual numbers. 240x decline in 18 months. Not "costs are falling" - a specific, sourced, cited number that changed how people think about the market.
02
Operator Brain in a VC Body
He ran News Feed ranking at Meta. He knows what it means when a recommendation algorithm gets even slightly better at retaining attention. That operational intuition shapes every investment thesis he writes.
03
Moats, Measured
His framework on AI-era defensibility - the idea that marketplace density and trust are structural, not labor-based - is a genuine contribution to how practitioners think about competitive advantages in software.
05
Range Without Noise
Studio Ghibli picks. ServiceTitan S-1 breakdowns. Reddit's data licensing model. The newsletter covers everything but never feels scattered - there's always a consistent analytical lens underneath the range.
06
12K Subscribers, Zero Viral Moments
The newsletter didn't explode from a tweet thread going viral. It grew from consistent usefulness. That's actually harder to do - and a better signal about the quality of the work.
Background

The Education Stack

Harvard Business School
MBA - Baker Scholar (Top 5% of graduating class)
2018 - 2020
Columbia University
BS in Computer Science - The technical foundation he brings to every analytical framework
~2010 - 2014
In His Own Words

The Quotes Worth Saving

The cost of intelligence is in free fall, with the cost of GPT-4 equivalent intelligence having fallen 240x in the last 18 months.
AI cannot fabricate real-time liquidity, courier density, reputation history, or a canonical identity graph. Marketplace density and trust are structural, not labor-based.
Startups are now pitching AI that doesn't just improve workflows but fully owns them - setting the stage for a transformative shift, moving the TAM from software budgets to include the entire labour spend.
AI compresses labor-based scale advantages in software and digital work. The question every software business needs to answer: what structural advantage do you have that AI can't compress?
Track Record

The Highlights

Fun Facts

Things You Didn't Know About Tanay

2008
The year @tanayj joined Twitter. He's been on the platform for over 16 years - an early adopter who built his 72K following through consistency, not virality.
240x
The AI inference cost decline he tracked over 18 months. That's not a trend - that's a revolution. He put the specific number on it before most people were even asking the question.
Top 5%
Baker Scholar status at Harvard Business School. Fewer than 1 in 20 graduates earn it. It's an academic distinction that actually signals how well someone thinks under pressure.
48 hrs
He takes spontaneous 48-hour trips to places like Iceland, documents them on Instagram. The same person writing about AI moats also photographs glaciers on a weekend. Makes sense.
Studio Ghibli
He writes about Studio Ghibli recommendations in the same newsletter where he breaks down AI agent economics. Range is a feature, not a bug.
Mumbai → NYC
Grew up in Mumbai, India. Moved through London and San Francisco before landing in New York. International perspective is baked into how he reads global tech markets.
Tanay's Newsletter

Weekly. Rigorous. Actually Useful.

12,000+ readers who want to understand AI, SaaS, and tech business models before everyone else does.

Subscribe on Substack
Share This Profile