.png)
Mercor, the startup founded in 2023 by three 22-year-old entrepreneurs, has just raised $350 million in a Series C round, bringing its valuation to $10 billion. In less than three years, this AI-powered recruitment platform has become a key player for artificial intelligence laboratories looking for human experts to train their models. With an annualized turnover of $500 million and more than 30,000 contract workers paid $1.5 million daily, Mercor embodies the new labor economy in the era of generative AI.
In October 2025, Mercor officially announced a $350 million Series C fundraiser, led by Felicis Ventures. This operation brings the valuation of the startup to 10 billion dollars, five times more than in February 2025 during its Series B where it was valued at 2 billion dollars. Investors also include Benchmark, General Catalyst, and newcomer Robinhood Ventures.
This rapid growth places Brendan Foody (CEO), Adarsh Hiremath (CTO), and Surya Midha (Chairman of the Board), all aged 22, in history as the youngest self-taught billionaires in tech, surpassing even Mark Zuckerberg who became a billionaire at 23.
The numbers speak for themselves: Mercor went from an annualized revenue (ARR) of $75 million in February 2025 to $100 million in March, and then to $500 million in September 2025. This trajectory makes it one of the fastest-growing startups in the tech sector, comparable to Anysphere's Cursor, which reached ARR 500 million around a year after launching its main product.
The startup now manages more than 30,000 contract workers around the world, to whom it collectively pays more than 1.5 million dollars per day. And despite this impressive scale, Mercor remains profitable — a rare feat in the world of hypergrowth startups.
Founded in 2023 by three former Thiel fellows who left Georgetown and Harvard, Mercor initially started as an automated recruiting platform using AI to connect freelance engineers with startups. The system automated the sorting of resumes, offered AI-driven interviews, and even managed payroll.
But the founders quickly identified a much more lucrative opportunity: AI labs like OpenAI, Anthropic, and Meta were desperate for human experts to train and refine their artificial intelligence models. These businesses needed scientists, doctors, lawyers, lawyers, bankers, and other highly skilled professionals who could provide specialized knowledge that machines alone could not capture.
Mercor's timing was perfect. In June 2025, Meta invested $14.3 billion to acquire 49% of Scale AI, Mercor's main competitor in data labeling and AI training. Alexandr Wang, the founder and then CEO of Scale AI, left his position to join Meta as part of this deal.
This acquisition raised major concerns about Scale AI's neutrality. Several major AI laboratories, including OpenAI and Google DeepMind, would have cut their ties with Scale AI following this investment, fearing conflicts of interest. As Adarsh Hiremath told Forbes, “It's not often in startups that your biggest competitor gets torpedoed overnight.”
This situation created a market vacuum that Mercor quickly filled, attracting these leading customers who were looking for reliable alternatives.
Mercor offers a comprehensive platform that allows AI laboratories and technology companies to:
According to external estimates, Mercor would charge employers around 30% in placement fees, although the company does not officially publish its rates.
Professionals register on the platform by completing a 20-minute interview driven by AI that assesses their skills and creates a profile. They can then be connected to full-time, part-time, or hourly opportunities.
Compensation can reach up to $200 per hour for the most qualified experts, including those with doctorates, law degrees, or Wall Street experience. India remains Mercor's biggest source of talent, followed by the United States, while Europe and South America are growing rapidly.
What sets Mercor apart from competitors like ZipRecruiter, Otta, or even Scale AI is its focus on high value-added expertise tasks rather than standardized data labeling work. The platform uses proprietary reinforcement learning algorithms to match experts to highly specialized AI training tasks with high precision.
At TechCrunch Disrupt 2025, Brendan Foody revealed a fascinating strategic dimension of Mercor's model. Instead of signing expensive contracts with companies to access their proprietary data, AI laboratories are now using former senior employees of these companies through Mercor to obtain their industrial knowledge.
“There's an argument that Goldman Sachs doesn't like the idea of having models that can automate their value chain,” Foody explained. “This is definitely changing the competitive dynamic, and that's part of the reason why laboratories need us. Their customers don't want to give them data to automate large portions of their value chains, so they need to hire contractors who have previously worked at these companies.”
Foody envisions a future where Mercor pays out tens of billions of dollars a day to contract workers, as machine training becomes a dominant work category. For him, Mercor is not eliminating jobs but identifying tasks that humans should perform in an AI-driven economy — jobs that AI cannot do, such as training AI models, managing complex decisions, or performing creative and strategic roles.
However, this vision is drawing criticism. For some observers, this model resembles a dystopian gig economy where skilled workers become precarious providers serving AI giants. Others are concerned that specialized knowledge from established industries “escapes through the back door” via former employees, to be ultimately used to automate these same sectors.
Mercor has faced some criticism regarding the authenticity of the positions offered and the use of AI-conducted interviews. A critical article published on Medium in 2025 raised questions about “fake jobs” and data collection. Although Mercor has published clear policies stating that it “will never use your interview data to train AI models” and that it will not sell interview or resume data to third parties, these concerns persist in some circles.
In September 2025, Scale AI sued Mercor for trade secret misappropriation. Scale AI alleges that a former employee who joined Mercor “stole over 100 confidential documents regarding Scale's customer strategies and other proprietary information.” This legal battle illustrates the intensity of competition in this booming sector.
Mercor's wealth is intrinsically linked to a few AI laboratories. Even Sam Altman, CEO of OpenAI, warned that we are in an AI bubble. If this bubble bursts or if the need for human training decreases with advances in autonomous AI, Mercor's model could be questioned.
In addition, major players such as Uber are positioning themselves in this market. Uber CEO Dara Khosrowshahi announced this week that the company would become a “work platform,” offering AI training missions. OpenAI has also launched its own recruitment platform, which could lead the AI giant to create its own reinforcement training service powered by human experts.
Mercor aims to extend its platform beyond the technology sector to specialized industries such as health and legal services, where expertise is essential and the stakes are the highest. The company is also developing new AI-powered recruitment markets designed to evolve the way experts are connected globally.
The new investments will focus on three priority areas:
Mercor also announced that it is developing more software infrastructures for reinforcement learning — a training method where the decisions of a model or agent are verified or contested, allowing it to incorporate feedback and improve over time.
To take the business to the next level, Mercor recently appointed Sundeep Jain, a former product director at Uber with decades of experience, as its first president. This strategic hire brings mature leadership experience to a team of very young but extraordinarily ambitious founders.
Mercor's meteoric rise epitomizes the promises and tensions of the AI economy. On the one hand, the startup offers substantial income opportunities to qualified professionals around the world, democratizing access to well-paid intellectual work. On the other hand, it raises fundamental questions about the future of work, intellectual property, and the role of humans in an increasingly automated world.
As Foody himself admitted: “There were some things we did very well, but there were a lot of things where we were lucky. Like being born at the right time, being perfectly positioned to seize one of the most exciting market vacuums in history.”
It remains to be seen whether this model is sustainable in the long term or if it simply represents a transitory phase in the evolution of artificial intelligence. One thing is certain: Mercor has successfully positioned itself at the heart of a major transformation in the global economy, and its journey will continue to be closely watched by investors, workers, and industry analysts.
Sources:
