▶ Watch ↗AI Engineer World's Fair 202618:44
The Loop Is the Product — Roland Gavrilescu, Introspection
Read the full talk →Key ideas
Scroll to read ↓Roland Gavrilescu’s blueprint for improving agents connects execution, human judgment and production experiments: capture what a system learns in portable recipes, then optimize for work users value at a cost that makes sense.
- Build the feedback path as carefully as the worker: signals guide execution, verifiers check success, and artifacts inform the next improvement loop.3:48 ↗
- Preserve operating knowledge in versioned agent recipes: failures become evaluations, repeated behavior becomes skills, and human judgment remains part of the system.4:55 ↗
- Let humans calibrate the desired judgment and agents implement the evaluations. Then test whether users value that judgment before promoting a recipe change.14:58 ↗
- Establish that the output is valuable, then make its economics viable. Efficient execution only becomes a useful product advantage when customers want the work.17:27 ↗