Mercor

Mercor is an AI data and workforce platform founded in 2023 by Brendan Foody and Adarsh Hiremath, headquartered in San Francisco with offices in New York City and London. The platform connects millions of domain experts with leading AI labs and enterprises, enabling high-value human contributions to frontier model training, evaluation, and enterprise AI deployment. According to the company, experts on the platform are paid over $4 million per day collectively, and the platform has facilitated hiring of over 100,000 contractors.

What It Is

Mercor sits at the intersection of human expertise and AI development, operating as a marketplace and infrastructure layer that recruits, vets, and deploys domain experts—from medical professionals and PhDs to finance and cybersecurity specialists—to perform AI training tasks, data annotation, and model evaluation. The platform also offers enterprise services that deploy custom AI agents trained on company-specific knowledge and workflows, and publishes the APEX benchmark family to measure AI's real-world impact on professional work.

How the Expert Network Works

Mercor uses AI-powered, role-specific interviews to dynamically vet candidates across 300+ professional fields. Experts apply to roles listed on the platform, are screened through automated interviews, and—once approved—perform tasks such as ranking model outputs, writing ideal completions, and providing domain-specific feedback that feeds directly into RLHF (Reinforcement Learning with Human Feedback) pipelines. The platform supports daily payouts and lists roles spanning medical experts, physics PhDs, management consultants, cybersecurity professionals, and more.

Enterprise AI Services

Mercor Enterprise brings the same expert network and evaluation infrastructure to Fortune 500 companies. The service is structured around four stages:

• Agent Diagnostics – Capturing workflows and tribal knowledge to identify high-value automation opportunities
• Agent Deployment – Building and deploying custom AI agents trained on company-specific knowledge
• Agent Benchmarking – Independent, repeatable benchmarks powered by the expert talent network to pressure-test AI products
• Data Monetization – Helping enterprises safely license their proprietary data

The company states that its own internal deployment of these systems resolved 90% of support tickets end-to-end by AI and saved $1.4M in customer support automation.

APEX Benchmark and Research

Mercor publishes the APEX (AI Productivity Index) benchmark family, which measures whether AI models can perform economically valuable professional work. The APEX-Agents benchmark has been covered by TechCrunch as a measure of AI agent readiness for the workplace. This positions Mercor not just as a data supplier but as an evaluator of AI capabilities in real-world professional contexts.

Recognition and Backing

According to the company's mission page, Mercor has been recognized on Forbes AI 50, Forbes Cloud 100, LinkedIn Top Startups, Bloomberg's 24 AI Startups to Watch, and Newsweek's America's Greatest Startup Workplaces. The company reports a $10B valuation and $2B+ revenue run rate. Investors include Benchmark, General Catalyst, Felicis, Robinhood Ventures, and Menlo Ventures. In July 2026, Fortune reported that Mercor acquired Deeptune to scale reinforcement learning environments.