Judgment Labs provides a platform that evaluates long reasoning traces, tool use, and memory to surface failure patterns in agent trajectories and convert production interactions into concrete fixes. Its product is already in production at multiple agent-native companies, where it powers monitoring and improvement cycles for deployed agents. The company positions itself as the continuous-improvement layer for deep agents, addressing evaluation challenges that arise as models execute multi-step tasks. Judgment plans to use its recent funding to hire AI researchers and engineers in San Francisco and to expand its forward-deployed engineering team serving customers. The company is backed by Lightspeed Venture Partners, Nova Global, SV Angel, Valor Equity Partners, and Dynamic.

Total raised

$32M

Funding rounds

2

Latest round

Series A

Latest activity

May 2026