How AI is reshaping work and who gets to do it, according to Mercor's CEO

What It Covers

Mercor CEO Brendan Foodi explains how his company reached $10 billion valuation by connecting AI labs with former Goldman Sachs, McKinsey employees earning $95/hour to create training data that automates their former employers.

Key Questions Answered

• •Power law talent distribution: The top 10-20% of contractors drive majority of model improvement, similar to company performance dynamics. Mercor's competitive advantage comes from building proprietary systems to source, assess, and retain this elite segment through referral networks and performance data.
• •Knowledge work transformation: Companies will shift from redundant task execution to one-time agent training, similar to software's fixed cost model. Investment bankers will build evaluation datasets once rather than repeatedly analyze similar data rooms, fundamentally restructuring how professional services operate and scale.
• •Customer concentration strategy: Mercor grew from $1 million to nine figures in eleven months, reaching $500 million annualized revenue. Despite concentration risk with OpenAI, the model mirrors NVIDIA's hyperscaler concentration, betting enterprise transformation toward agent training represents massive untapped market expansion opportunity.
• •Contractor protection framework: Mercor uses contracts and guidance to prevent former employees from sharing proprietary company documents or trade secrets. This legal structure addresses corporate espionage concerns while enabling workers to leverage general industry knowledge and expertise accumulated through their professional experience.