Product Engineer, Agent
AI-firstOn-siteSeniorFull-time
claude codecodexopencode
.·-·
·#*:
::*#
#+-+
=#+-+#+-+**::*
-+#*::**::*#·-
*::*#·-·#·-·..
·#·-·.........
..#·-·#·-·**
·-·**::**::+
**::*#+-+#+-
-+#+-+======
1 open role
Judgment builds learning infrastructure for AI agents in production, ingesting traces, tool calls, and outcomes to generate structured signals (failure modes, rubrics, evals) that help teams validate and ship agent improvements. Its product includes a 'Judgment Agent' for large-scale trace investigation, verification tooling for agent changes, and an SDK for integrating with tools like Claude Code, Codex, and OpenCode.