Staff Software Engineer, Agent Eval Platform

📋 Description

• Role: implement the evaluation signal for agent judging
• Work on the runtime that executes multi-turn agent scenarios
• Build observability and tracing for the agent platform
• Develop stateful simulation environments used in runs
• Aim for self-serve evals across teams with reliability
• Contribute to a scalable evaluation harness and data pipelines

🎯 Requirements

• 8+ years building production backend or infrastructure
• Strong in Python or Go (ideally both)
• Experience designing systems with real traffic at scale
• Comfort turning fuzzy agent behavior into measurable signals
• Open to ambiguity and novel problems without textbooks
• Proficient in distributed systems and async programming