Agent evaluation set
Definition
Agent evaluation set is an agent evaluation set is a curated collection of example inputs and expected behaviours used to test whether an agent still performs acceptably after prompt or model changes.
Evals are the regression tests of agent ops. Run them before raising autonomy. Anecdotal “it looked fine” demos are not quality control.
Go deeper
See what your booked-meeting rate could be.
Book a walkthrough More terms
All 624 terms