01What it sells and who buys it
Mercor evaluates and matches experts to AI training and assessment work for labs and enterprises. It expanded from developer recruitment into specialized human work supporting frontier AI.
02How the business makes money
Enterprise customers pay for expert work and project delivery. Worker compensation and customer billings are distinct; Mercor commercializes matching, staffing and quality management.
03Marketing and customer acquisition
Mercor built both talent supply and nearby demand before expanding into AI research work.
- 01
Connect both sides of an initial market
SourceThe founders worked with an IIT Kharagpur developer group and connected talent to friends’ companies and their own projects, creating actual demand alongside supply.
- 02
How distribution and adoption expanded
SourceThe founders worked with an IIT Kharagpur developer group and matched people to their own projects and friends' companies. That established demand alongside talent supply. Large expert requests from AI labs then expanded the network and provided evidence of staffing and evaluation capacity.
- 03
Financing the next stage
SourceIn October 2025, Mercor raised a $350 million Series C led by Felicis, with Benchmark, General Catalyst and Robinhood Ventures participating. The round valued it at $10 billion as demand for expert AI work expanded.
How the growth mechanism works
- Recruit specialists
- Assess them with AI interviews
- Assign research projects
- Earn repeat orders through delivery
Research-lab demand expanded toward expert training data. Mercor applied its assessment and staffing capabilities to that need. Growth requires both client acquisition and a reliable specialist supply.
04Funding and expansion
In October 2025, Mercor raised a $350 million Series C led by Felicis, with Benchmark, General Catalyst and Robinhood Ventures participating. The round valued it at $10 billion as demand for expert AI work expanded.
More about the founding and key events
Brendan Foody, Adarsh Hiremath and Surya Midha started Mercor in a college dorm room in 2023. Its first product used AI interviews to match applicants with employers. A January 2024 announcement reported 100,000 people across 25 countries and seven-figure ARR before fundraising.
Foody described working with a developer group at IIT Kharagpur to find people, then connecting them with his own projects and friends’ companies. The team developed nearby demand and talent supply together. It later expanded to requests from AI labs.
Those labs needed many experts at short notice. Foody described a request for 300 people in two days. In ordinary hiring, performance can take months to evaluate. For model training, the quality of an expert’s output could be assessed in days. Repeated sourcing, evaluation and placement improved both matching and delivery.
As model companies began buying expert knowledge at scale, Mercor’s existing network and evaluation system met that demand. It reached a $10 billion valuation in 2025. In 2026, the three 22-year-old founders were each estimated to be worth $2.2 billion. Its announced Deeptune acquisition extended the business into environments where AI can practice work.