Product Manager, Claude Code Model Performance

Overview

Product Manager for Claude Code’s model performance team, driving end-to-end model launches and shipping developer-facing improvements.

What you'll do

• Own Claude Code model launch planning and execution using readiness criteria.
• Design and implement agentic evals to measure real-world coding performance.
• Drive the engineering team’s eval roadmap.
• Partner with researchers on coding capabilities to define target behaviors and influence development with evidence.
• Talk with users and analyze transcripts to identify capability gaps.
• Synthesize internal and external signals into clear product priorities.

What you'll need

• Personally built agentic evals (e.g., SWE-bench-style task suites).
• Daily Claude Code user who can articulate behavior changes.
• Engineering background plus 2+ years in product management, or equivalent experience driving product direction.
• Deep grasp of AI concepts, comfortable working on model behavior, prompt engineering, and evaluation methodology.
• Systems-thinking approach: build infrastructure to prevent whole classes of problems.
• Comfort launching products/capabilities in ambiguous, research-adjacent environments.
• Bachelor’s degree or equivalent combination of education, training, and/or experience.

Details

• Location: San Francisco and Seattle only.
• Work mode: location-based hybrid policy; expect staff in offices at least 25% of the time.