Don't Trust Your AI Agents. Test Them Like a Scientist — Stephanie Jarmak, Omni Analytics
AI Engineer — Oct 23, 2026 — talk
Why watch
Stephanie Jarmak explains how to properly evaluate and test AI agents like a scientist. She covers why testing harness changes can swing results by 30%, when to trust vs question public benchmarks, how to design effective evals with realistic tasks and repeated runs, and implementing quality gates that actually stop poor agent outputs. Essential viewing for engineers building AI agents who need systematic testing approaches.