Robotics Model Evaluation Lead

Dyna Robotics; Redwood City (CA); On-site; USD 150,000 - 230,000; Full time

Job summary

Dyna Robotics is seeking a Model Evaluation Lead to head the team that assesses how our robot policies perform and whether new model versions improve capabilities. You will manage, plan, and execute evaluations, working closely with researchers to define tests, scores, and learnings.

This hands-on leadership role blends people management, experimental rigor, and technical judgment to ensure results are trustworthy and actionable for deployment decisions.

Qualifications

• Experience in robotics, ML, applied research, or model evaluation with ownership of evaluation design and analysis.
• Demonstrated experience managing and developing a technical team, including setting priorities and giving feedback.
• Strong experimental judgment: define success, control variables, identify gaps, assess meaningful improvements.

Responsibilities

• Manage workflows and priorities for the evaluation team.
• Collaborate with researchers to translate questions into evaluation plans and hypotheses.
• Design task definitions, scoring rubrics, and test scenarios reflecting real operating conditions.
• Make results repeatable and trustworthy with standardized setups and logs.
• Turn failures into actionable findings by distinguishing model vs hardware or setup issues.
• Own regression coverage and reporting across tasks and model versions.
• Improve evaluation tools and workflows with automation and dashboards.

Skills

Robotics experience

Machine learning

Team leadership

Experimental judgment

Technical writing

Communication

Tools

Python

Job description

Dyna Robotics is seeking a Model Evaluation Lead to head the team that assesses how our robot policies perform and whether new model versions improve capabilities. You will manage, plan, and execute evaluations, working closely with researchers to define tests, scores, and learnings.

This hands-on leadership role blends people management, experimental rigor, and technical judgment to ensure results are trustworthy and actionable for deployment decisions.