- Mercor hit $500 million in annualized revenue in August 2025 — just 17 months after crossing $1 million — making it one of the fastest revenue ramps in startup history.
- Brendan Foody dropped out of Georgetown at 20, received a Thiel Fellowship, and co-founded Mercor with two high school debate teammates from Bellarmine Prep.
- The company raised $350 million in its Series C at a $10 billion valuation in October 2025, making its three founders the youngest self-made billionaires in the world.
- Mercor now employs over 30,000 domain experts — doctors, lawyers, scientists — who train AI models for OpenAI, Google DeepMind, Meta, and Amazon.
30,000 Experts, $10 Billion, and a CEO Who Just Turned 23
Mercor is the company that teaches AI what only humans know. Its platform matches domain experts — consultants, physicists, constitutional lawyers — with AI labs that need human judgment to train and evaluate their models. Clients include OpenAI, Google DeepMind, Meta, Amazon, and Nvidia. Contractors earn an average of $95 per hour to test and score model performance, writing the rubrics and evaluations that shape how frontier AI reasons.
The company was founded in 2023 by three college dropouts. Brendan Foody is the CEO. He runs the fastest-growing company in history to cross the $1 million-to-$500 million revenue threshold — and he did it before his 23rd birthday. To understand how, you have to go back to a middle school parking lot in San Jose.
A Kid Who Sold Donuts Before He Could Drive
Foody grew up in the Bay Area and attended Bellarmine College Preparatory, a Jesuit school in San Jose known for its nationally ranked speech and debate program. That’s where he met Adarsh Hiremath and Surya Midha — the two co-founders who would later build Mercor with him.
But the entrepreneurial instinct came earlier. In eighth grade, Foody launched what he called Donut Dynasty. He biked to Safeway, bought a dozen donuts for $5, and sold them to classmates for $2 each after school. When the principal shut him down for selling food on campus, he moved his stand fifty feet off school property to comply. He paid his mother $20 a week to drive him in her minivan.
”When I was in college, work was something I had to be disciplined to do. When I started Mercor, it became something I couldn’t stop thinking about.” — Brendan Foody
Georgetown, Debate, and the Decision to Drop Out
Foody enrolled at Georgetown University in 2021, studying economics and business administration. Hiremath went to Harvard. Midha went to Stanford. All three kept in touch, united by the same restlessness that had driven them through debate tournaments.
By sophomore year, the restlessness won. In spring 2023, while his classmates crammed for finals, Foody was already testing a new theory of work. He dropped out before exams. All three founders received Thiel Fellowships — $100,000 grants that require recipients to leave college and build something instead.
A Hackathon in São Paulo Changed Everything
The catalyst was a hackathon in São Paulo, Brazil. Foody, Hiremath, and Midha stumbled onto a model that would define their company: match companies with skilled engineers abroad, handle the logistics, and take a small cut of each contract. Within nine months, that model generated $1 million in revenue.
But the real insight came as AI labs began spending billions on post-training — the process of refining models after initial pre-training. Labs needed human experts to write evaluations, create rubrics, and test whether a model’s reasoning held up in specialized domains. Finding those experts at scale was nearly impossible.
”Everyone’s been focused on what models can do. But the real opportunity is teaching them what only humans know — judgment, nuance, and taste.” — Brendan Foody
Foody saw that the hiring problem and the AI training problem were the same problem. Mercor pivoted from general recruiting to building the infrastructure that connects domain experts with AI labs.
From $35 Million to $500 Million in Eight Months
The numbers tell a story that reads like a typo. At the end of 2024, Mercor’s annualized revenue was $35 million. By August 2025, it crossed $500 million — a greater than 10x increase in eight months. The company’s AI-driven evaluation system conducts 20-minute video interviews, generates transcripts, and semantically matches candidates to client briefs written in natural language.
In February 2025, Mercor raised a $100 million Series B led by Felicis at a $2 billion valuation — eight times its previous $250 million valuation. Eight months later, it quintupled again. The $350 million Series C closed in October 2025 at $10 billion, with Benchmark, General Catalyst, and Robinhood Ventures participating. The average age on the team was 22.
Three 22-Year-Old Billionaires and a Company That Keeps Hiring
The $10 billion valuation made Foody, Hiremath, and Midha the youngest self-made billionaires in history — younger than Mark Zuckerberg was when Facebook hit comparable milestones. Bloomberg named Mercor one of its 24 AI Startups to Watch in 2026. The company hired the former head of Human Data Operations at OpenAI and the previous head of Growth at Scale AI.
An outsized share of revenue comes from a handful of AI giants, a concentration risk that Foody has acknowledged. But the bet is that demand for expert human evaluation will only grow as models get more capable and the stakes of their outputs rise.
”We’ll automate maybe two-thirds of knowledge work. And that’ll be incredible, because it lets us do things like cure cancer and go to Mars.” — Brendan Foody
What Comes After Training the World’s Smartest Models
Foody hasn’t taken a day off in three years. The company that started as a recruiting hack at a Brazilian hackathon now sits at the center of the AI supply chain — the layer between human expertise and machine intelligence. Mercor’s next move is expanding beyond AI labs into enterprise hiring, applying the same AI interview and matching technology to traditional recruitment at scale.
The irony is hard to miss. A company built to help AI learn from humans may end up reshaping how humans find work entirely. Foody is 23. Mercor is three years old. And the market for teaching machines what they cannot learn on their own is just getting started.
