Mercor, now valued at $2bn, automates staffing by using AI interviews and models to screen candidates and match them with jobs, with over 300,000 candidates assessed.
Brendan Foody says Mercor's defensibility comes from marketplace network effects and a usage data flywheel built from more than 1,000 placed roles, and he calls the labor market the largest and most inefficient market in the world.
Foody says competitors built services businesses rather than products, while Mercor handles payments and international compliance to gain data across the whole transaction and cut reliance on recruiters.
Mercor Replaces Manual Staffing With AI Interviews and Matching
Mercor is betting that hiring can run as software instead of as a recruiting service. Foody calls the labor market the largest and most inefficient in the world, and manual processes limit how well it works. Each placement feeds data that sharpens the models at predicting who will perform, which rivals built on services lack.
Takeaways
Mercor automates the manual work of staffing. Mercor automates resume review and AI interviewing, then matches candidates with jobs. It grew from zero to mid seven-figure revenue within a year, and more than 300,000 candidates have gone through its assessment. Employers search the talent pool using AI-powered evaluation.
Rivals built services businesses, not products. Foody says Toptal began as an arbitrage on Eastern European developers with fees often averaging 40 percent, and calls Turing a manual business with heavy AI marketing. Several former Turing executives now advise Mercor.
Predicting job performance is the core challenge. Mercor uses fine-tuned language models and learn-to-rank models, with features from GitHub contributions to interview audio and video. Audio carries more signal than transcripts, and even camera quality adds some. The models run at the scale of hundreds of thousands of interviews.
Defensibility comes from network effects and data. More than 1,000 placed roles teach the models to predict performance at the next role. Foody says simply operating like LinkedIn would miss that data. Handling payments and international compliance gives Mercor data across the whole transaction.
Engineering data is the base for other roles. Mercor has hired well over 1,000 software engineers, mostly in India, and is expanding into business, medical and legal roles. Transfer works best between similar roles such as full stack and backend, and struggles with designers. Customer preference data is the main validation.
Transparency builds trust with clients and candidates. Clients see recordings of AI interviews and summaries of why candidates were chosen. Candidates are evaluated fairly and get free scores and written feedback, while interview feedback is still in beta and career coaching is being explored.
Global hiring and fewer recruiters cut costs. Foody says reducing reliance on recruiters is a major saving, and employers hire talent in places such as India and South America. International talent is less dense than Silicon Valley's, but adding one or two contractors per US person is high leverage.
How Mercor began
Foody ran a high school cloud consulting business that charged startups $995 to help them claim AWS Activate credits worth about $25,000.
Tap a name to go there · tap a line to see why they are linked
On this episode 3
Show 1
He met cofounder Adarsh Suriya at Bellarmine San Jose, where their debate team won the largest national policy tournaments.
Foody launched Ceros, a browser-based virtual machine company.
The founders needed developers and hired members of a code club in India, who were getting big tech offers at about $5 an hour.
They went full time in summer 2023, bootstrapped to a seven-figure run rate, raised a small General Catalyst round and dropped out at 20.