Mercor co-founder and co-CEO Adarsh Hiremath addresses the AI onboarding problem
In this episode
As the co-founder and co-CEO of Mercor, Adarsh Hiremath has built his career around helping enterprises deploy agents, and crucially, ensuring that those agents act as intended. In his conversation with 1Password CTO Nancy Wang and Google Gemini’s Dev Tagare, Adarsh explains why evals are key to successful agent deployments, what most teams get wrong in the production rollout process, and why agents fail without company-specific context.
How to build agents that do the right work
Correct outputs don’t prove an AI agent’s reasoning is sound
Evals are a core infrastructure investment and must be continuous
Agent failures in background systems are the hardest to catch and fix
Agents need company-specific knowledge to overcome the onboarding gap
AI will shift from being measured in tokens to being measured by outcomes
Also available in audio format on the following platforms: Apple Podcasts, Spotify
Adarsh Hiremath
Adarsh Hiremath is the co-founder and co-CEO of Mercor, an AI company providing the human expertise essential to AI development. Mercor has become a trusted partner to leading AI companies and built a network of tens of thousands of expert professionals who help train frontier AI models for top labs and enterprises, while building APEX benchmarks that measure AI capabilities across economically valuable roles. Founded in 2023, Mercor has become one of the fastest-growing companies in Silicon Valley history.
Get the episode transcript
More episodes
OpenAI Agent Security Lead Fotis Chantzis discusses one of the biggest unsolved problems in AI
Vercel Chief Product Officer Tom Occhino reflects on React and how AI transforms software creation
Cursor Head of Security Travis McPeak considers how to secure agents without slowing work
