Stanford eCorner (leadership/innovation)

Brendan Foody (Mercor) - Agentic Data and the Future of AI [Entire Talk]

Brendan Foody, co‑founder of Mercor, addressed Stanford’s entrepreneurial series, outlining how his recruiting startup transformed AI model training by matching elite professionals with enterprise projects. He traced Mercor’s meteoric rise from a modest consulting venture to a $1 billion annualized revenue company in less than two years, positioning it as the fastest‑growing startup in history and a primary data vendor for leading AI labs. The talk highlighted the industry’s pivot from low‑skill crowdsourced labeling to "agentic data," where highly skilled experts create structured evaluation environments and reward functions. This shift enables more accurate benchmarking and rapid reinforcement‑learning improvements, driving the next generation of AI agents. Foody also shared concrete metrics: Mercor scaled from a $1 million run‑rate to $1 billion in just 20 months, securing multi‑billion‑dollar valuations. Foody peppered his narrative with vivid anecdotes—a middle‑school donut resale that taught pricing strategy, a high‑school debate team that forged co‑founder chemistry, and a chance meeting with an OpenAI engineer that unlocked the agentic data market. He emphasized the dynamic division of labor among the founders, with each focusing on engineering, recruitment, or enterprise expansion as needs evolved. The implications are clear: mastering agentic data pipelines can accelerate AI capabilities and create massive market opportunities. Mercor’s model demonstrates how young entrepreneurs can leverage niche expertise, rapid iteration, and strategic partnerships to dominate emerging tech ecosystems, offering a blueprint for future AI‑centric ventures.