Talks

Upcoming

Nov 16–19, 2026 · AI Coding Summit, Jersey City NJ

Debugging the Agent That Refuses to Fail the Same Way Twice

Capture a failed agent run at the boundary, replay it deterministically without live LLM calls, and promote that trace into a regression test.

Details

Past

2026 · AI Engineer World’s Fair, San Francisco

FinOps for AI Agents: Who Spent All the Tokens?

Cost, attribution, and policy once agents start spending on their own.

Watch

2026 · AI Engineer World’s Fair, San Francisco

Your Agent Failed in Prod. Good Luck Reproducing It.

Record a failed run at the boundary, replay it without live LLM calls, and turn that trace into a regression test.

Watch