ZeroEval (YC S25)

A new way to measure the quality of AI agents using human feedback.

Our mission: build the engine that powers self-improving software.