He Missed IIT's Computer Science Cut By 0.01 Points
Then Built The Company Trying To Buy Google Chrome

Aravind Srinivas's Childhood In Chennai, India

Aravind Srinivas was born on June 7, 1994, in Chennai, in the southern Indian state of Tamil Nadu. He grew up in a middle-class household — his father worked as a financial accountant, and his mother worked for the central government.

From early on, his mother had one specific dream for him: getting into the Indian Institute of Technology, or IIT — the most competitive and prestigious engineering institution in India. Srinivas has recalled that whenever their bus passed the IIT Madras campus, his mother would point at it and say, "This is where you're going to study."

He delivered. After years of intense preparation — the kind millions of Indian students go through for the notoriously difficult IIT entrance exam — Srinivas cleared it and was admitted to IIT Madras for a dual degree (B.Tech and M.Tech) in Electrical Engineering.

The Tiny Setback That Made Aravind Srinivas Self-Taught

Once at IIT Madras, Srinivas tried to switch his major from Electrical Engineering to Computer Science after his first semester. He fell short — by just 0.01 of a percentage point.

Rather than treating the rejection as a dead end, Srinivas taught himself the fundamentals of programming and artificial intelligence on his own, outside of his official coursework. It worked far better than anyone could have predicted: while still a student at IIT Madras, he was already publishing research papers at top-tier AI conferences like ICLR and NeurIPS — venues that many working AI researchers never manage to publish in.

He graduated from IIT Madras in 2017 with a strong academic record, powered largely by knowledge he had built for himself after being told "no."

Aravind Srinivas's PhD At UC Berkeley — After Being Rejected By Stanford

After IIT Madras, Srinivas set his sights on a PhD in the United States. He didn't get into Stanford — a fact he has openly admitted in interviews, including one hosted by Stanford's own Graduate School of Business, where he laughed and said plainly, "I didn't get in here. That's the truth."

Instead, he enrolled at UC Berkeley, pursuing a PhD in Computer Science from 2017 to 2021 under the mentorship of Pieter Abbeel, a well-known figure in AI and robotics research. His academic focus was on reinforcement learning, contrastive learning for computer vision, and transformers for image generation — deep, technical territory that would later become directly useful in building Perplexity.

Aravind Srinivas: One Of The Few People To Work At OpenAI, Google Brain, And DeepMind

Between 2019 and 2022, while completing his PhD, Srinivas took on research internships at three of the most influential AI labs on the planet — OpenAI, Google Brain, and DeepMind. Very few researchers in the world have worked inside all three. At OpenAI, he contributed as a research scientist to the DALL-E 2 project, one of the earliest breakthrough AI image generators.

It was during a DeepMind internship in London that one of the defining moments of his career happened — not in a lab, but in a library. His rental apartment was uncomfortable, so he ended up sleeping at the DeepMind office some nights. The office had a library, and there he found a book called "In the Plex," about the founding of Google.

He read the book multiple times. The realization was simple but powerful: if two PhD students could build something as world-changing as Google, maybe he could build something too, instead of staying on the traditional academic path he had originally planned.

Founding Perplexity AI: Aravind Srinivas Leaves OpenAI Weeks Before ChatGPT

In August 2022, Srinivas co-founded Perplexity AI together with three collaborators: Denis Yarats (a former Meta AI researcher, who became CTO), Johnny Ho (a former Quora engineer, who became Chief Strategy Officer), and Andy Konwinski (co-founder of Databricks). In a detail that's easy to miss, Srinivas left his research role at OpenAI just weeks before ChatGPT was released to the public in November 2022.

The team's core idea was straightforward but ambitious: instead of giving users a long list of blue links like Google, or a chatbot that sometimes made things up like early ChatGPT, Perplexity would give direct, conversational answers backed by real-time web search and visible citations for every claim.

Srinivas has said his research background shaped the product directly: because he had spent years having to cite sources properly in academic papers, it felt natural to build citations into Perplexity's answers from day one — long before "AI hallucination" became a household concern.

Perplexity AI's Valuation: From $121 Million To $20 Billion

Perplexity's growth in valuation is one of the fastest ever recorded in the AI industry. In under three years, the company's value multiplied by more than 150 times.

The growth didn't happen in one jump — it climbed in a series of increasingly large rounds: past $500 million in January 2024, past $1 billion by April 2024, $3 billion by mid-2024, $9 billion by December 2024, $14 billion by mid-2025, and $20 billion by September 2025. Total funding raised across all rounds passed $1.5 billion, from backers including Nvidia, Jeff Bezos, SoftBank Vision Fund 2, Accel, and IVP.

How Aravind Srinivas Convinced Google's Own AI Chief To Invest In A Google Rival

One story captures Srinivas's fundraising hustle better than any pitch deck could: to secure investment from Yann LeCun — Meta's Chief AI Scientist and one of the most respected researchers in the field — Srinivas reportedly camped outside LeCun's office until he got a meeting.

It worked. Perplexity's investor list grew to include not just venture capital firms, but a striking number of personal checks from some of the most recognizable names in tech and business: Jeff Bezos, Nvidia (the company, as a strategic investor), Shopify CEO Tobi Lütke, former YouTube CEO Susan Wojcicki, and Figma CEO Dylan Field, among others.

Perplexity AI's $34.5 Billion Bid To Buy Google Chrome

In August 2025, Perplexity AI did something that made headlines around the world: it submitted an unsolicited $34.5 billion bid to acquire Google's Chrome browser — a browser used by billions of people every day.

Google did not accept the offer. But the bid itself, from a three-year-old startup, was a statement in itself: Perplexity was signaling that it saw browsers, not just search boxes, as the next real battleground for AI companies. Around the same time, Perplexity launched its own AI-powered browser, Comet, making it free worldwide by October 2025.

Growth like that did not come without controversy. Publishers accused Perplexity of scraping their content without permission, and Cloudflare publicly confirmed that Perplexity had used "stealth crawlers" to bypass website rules meant to block automated scraping. Perplexity responded by launching a Publisher Program — a revenue-sharing partnership with major outlets including Time, Fortune, Der Spiegel, Gannett, The Independent, and the LA Times, aiming to make sure journalism organizations get paid when their content powers Perplexity's answers.

Aravind Srinivas's Net Worth And His "80% Rule" For Building Products

By October 2025, Srinivas debuted on the M3M Hurun India Rich List as India's youngest self-made billionaire, with an estimated net worth of roughly $2.5 billion — almost entirely tied to his equity stake in Perplexity AI.

Beyond the money, Srinivas has become known for a specific product philosophy he calls the "80% rule," which he's discussed publicly, including at UC Berkeley's Haas Dean's Speaker Series: don't launch a product until it's genuinely useful, not just technically working. In his own words, launching something that's only 60% "perfect" is essentially launching something broken — users will try it, get frustrated that it doesn't solve their real problem, and simply never come back.

Aravind Srinivas And Perplexity AI: The Full Timeline, Year By Year

Five Business Lessons From Aravind Srinivas And The Perplexity AI Story

• A tiny rejection can push you toward your real strength

  Missing a computer science switch by 0.01 points forced Srinivas to teach himself — and that self-taught foundation became the exact skill set that later built Perplexity.

• Not getting into your first-choice school isn't the end of the story

  Rejected by Stanford, Srinivas built an equally strong career through UC Berkeley — proof that the specific institution matters less than what you do once you're inside one.

• One book, read at the right moment, can change your entire path

  A late night in a DeepMind office library, reading about how two students built Google, was enough to shift Srinivas from a planned academic career toward founding a company.

• Fundraising sometimes means literally showing up

  Camping outside Yann LeCun's office to get a meeting shows that persistence and simply being present can open doors that cold emails never will.

• Don't launch until it actually works

  Srinivas's "80% rule" — refusing to ship anything less than genuinely useful — helped Perplexity avoid the trap of shipping a technically-working but practically-useless first version.

Frequently Asked Questions About Aravind Srinivas And Perplexity AI