Mercor
Mercor connects human expertise with leading AI labs and enterprises to train frontier AI models. The company unlocks human potential in the AI economy by matching elite talent with AI development work, enabling professionals to teach machines judgment, nuance, and taste.
AI Tools by Mercor
(1)Mercor
AI Data Expert Marketplace
Discussions
No discussions yet
Be the first to start a discussion about Mercor
Products & Services
AI-driven platform (team.mercor.com) that sources, assesses, hires, and pays specialized workers. Features automated resume screening, candidate matching, and payroll management. Uses AI avatars to conduct 20-minute video interviews that evaluate skills and create candidate profiles. Includes automated crawlers that pull information from GitHub, LinkedIn, academic records, and personal portfolios. Offers deep semantic natural language search for employers to find candidates by describing roles. Companies can watch AI interviews and add candidates with one click, managing payments through the platform.
Platform connecting AI labs with domain experts for training frontier AI models through Reinforcement Learning from Human Feedback (RLHF). Provides high-fidelity training data and model evaluations from a network of 300,000+ credentialed professionals including PhDs, consultants, lawyers, doctors, scientists, and engineers. Experts are paid $60-$200+ per hour (average $95/hour). Platform can source hundreds of experts within 48 hours and includes API integrations for AI research workflows.
First-of-its-kind benchmark that evaluates AI models based on their ability to perform economically valuable knowledge work. Version 1.0 includes 200 expert-generated cases across investment banking, law, consulting, and medical practice using rubric-based grading. Designed to bridge the gap between AI evaluations and real-world professional tasks, measuring productivity impact and guiding next-generation model development.
Benchmark designed to test how effectively AI agents complete real, long-horizon tasks in professional fields like investment banking, consulting, and corporate law. Measures 'client-ready' performance by simulating complex workplace contexts with multiple documents and chat threads. Developed with input from experts at Goldman Sachs, McKinsey, Box, and Harvey AI. Open source and available on Hugging Face with infrastructure on GitHub. Initial findings show frontier models complete less than 25% of tasks, rising to only 40% after eight attempts.
Market Position
Mercor positions itself as the premium architect of the reinforcement learning economy and 'enduring AI infrastructure.' Key differentiators: (1) Moves up the value chain from low-skilled data labeling to high-skilled expert evaluations requiring domain expertise; (2) Differentiates from Scale AI by focusing on expert-level complexity rather than commodity labeling; (3) Differentiates from Toptal by using automated, tech-first AI vetting rather than slow human screening; (4) Competitive advantage built on speed (experts within 48 hours), scalability (300K+ network), and meritocracy (objective performance data over credentials). Uses 'AI-to-AI-to-Human' workflow. Fastest growing company in Silicon Valley history (grew 6400% year-over-year). More revenue per employee than world's largest tech/finance firms. Faces competition from Scale AI, Surge AI, micro1, Toptal, and potential internal hiring platforms from clients. September 2025 lawsuit from Scale AI alleging trade secret theft indicates intense competitive rivalry.
Founding Story
Mercor was founded in January 2023 by three childhood friends who met through high school debate competitions. Adarsh Hiremath and Surya Midha first met at age 10 while competing in elementary school debate tournaments, and all three competed together on their high school debate team. All three founders are Thiel Fellows who dropped out of college (Harvard and Georgetown) to pursue the venture. The company originally started with the mission of matching engineers in India with U.S. companies looking for freelance coders. They built an AI-powered recruiting platform using AI avatars to conduct interviews. This led them into the data labeling industry, where they began pairing expert-level contractors including lawyers, PhDs, consultants, and doctors with AI labs like OpenAI to train models. The company began 2024 operating out of an apartment with no US employees and only seed-stage companies as customers, but rapidly scaled to become one of the fastest-growing companies in Silicon Valley history.
Leadership
Founders
Brendan Foody
CEO and co-founder. 22 years old (as of 2025), Bay Area native, son of software engineers. His mother worked for Meta's real estate team and his father founded a graphics interface company. At age 16, he started a business charging friends for promotions on Amazon Web Services. Harvard dropout, Thiel Fellow, Georgetown University attendee (economics major). Met co-founders through high school debate team. Previously built a cloud services consultancy with Adarsh Hiremath. Featured on Forbes 2025 Under 30 list. Became one of the world's youngest self-made billionaires at age 22.
Adarsh Hiremath
CTO and co-founder. 22 years old (as of 2025), Indian-American, Bay Area native, son of software engineers. Attended Harvard University for two years before dropping out after sophomore year. Thiel Fellow. Conducted labor market research for Larry Summers while at Harvard. Met Surya Midha at age 10 competing in elementary school debate tournaments, later met Brendan Foody in high school debate. Previously built a cloud services consultancy with Brendan Foody. Featured on Forbes 2025 Under 30 list. Became one of the world's youngest self-made billionaires at age 22.
Surya Midha
COO and Board Chairman. 22 years old (as of 2025), Indian-American, Bay Area native, son of software engineers. Harvard dropout and Thiel Fellow. Initially pursued international relations at Georgetown University. Met Adarsh Hiremath at age 10 competing in elementary school debate tournaments, later met Brendan Foody in high school debate. Background in removing friction for global staffing. Featured on Forbes 2025 Under 30 list. Became one of the world's youngest self-made billionaires at age 22.
Executive Team
Brendan Foody
Co-Founder and CEO
22-year-old Bay Area native, Harvard/Georgetown dropout, Thiel Fellow, Forbes Under 30. Previously built cloud services consultancy. Son of software engineers, started AWS promotions business at 16.
Adarsh Hiremath
Co-Founder and CTO
22-year-old Indian-American, Harvard dropout (2 years), Thiel Fellow, Forbes Under 30. Conducted labor market research for Larry Summers. Previously built cloud services consultancy with Foody.
Board of Directors
Business Model
Revenue Model
Commission-based and cost-plus model. For direct talent placements, charges 30% recruiting/placement fee. For AI training and expert tasks, uses cost-plus hourly margin with a fixed percentage added to hourly rates paid by AI labs. Generates revenue through hourly finders' fees for contractor work. Pays contractors $60-200+ per hour (average $95/hour) and adds margin on top.
Pricing Tiers
One-time fee for direct recruiting and permanent placements
Fixed percentage margin on hourly rates ranging from $60-200+ per hour (average $95/hour paid to experts)
Target Markets
- AI research labs and frontier model developers
- Leading technology companies (Magnificent Seven)
- Enterprise AI development teams
- High-growth technology companies
- Investment banking and financial services
- Legal and corporate law firms
- AI model training and reinforcement learning
- Expert data labeling and annotation
- AI model evaluation and testing
- Domain expert consultation for AI development
- Software engineer recruitment and placement
- Technical talent hiring and vetting
- OpenAI
- Google DeepMind
- Anthropic
- Meta
History & Milestones
Released APEX-Agents benchmark for testing AI agents on long-horizon professional tasks
Reached $750 million recurring revenue; engaged in aggressive recruitment of competitors' employees with $500K-$2M signing bonuses
Faced lawsuit from Scale AI alleging trade secret theft via former employee Eugene Ling
Reached $500 million ARR (grew from $1M to $500M in 17 months)
Released AI Productivity Index (APEX) benchmark v1.0 for evaluating AI models on real-world knowledge work