Search architectures built for humans don’t work for AI agents.
At Localhost 2026, Nitya Sridhar from
@ExaAILabs went behind the scenes on what it takes to build high-performance agent search at scale: using the Generator–Verifier pattern with lower-cost subagents to improve accuracy without letting costs climb.
Watch the full conversation to learn how to build agent search systems that remain accurate, efficient, and cost-effective as they scale.
Full episode: