Robotics Research

About the Role

You will define what good looks like for the next generation of physical AI. That means designing the evaluations the field is missing, running them across our models and the public ones, and figuring out what the data is telling us. Your work shapes what the rest of the team builds.

What You'll Do

• Lead our evaluation and benchmarking research. Run the existing benchmarks from industry and research, and design new ones where they fall short.
• Investigate how models behave on real robots versus in simulation. Close the gap.
• Drive research into intervention, autonomy, and the boundary between what a model can do alone and what still needs a human.
• Publish the work where it makes sense. Set the academic and public reputation of the team.
• Work closely with the research engineers to take ideas from paper to deployment.

What We're Looking For

• A strong record in evaluation, benchmarking, or applied robotics research. Public work we can read is a plus.
• Comfortable on real hardware, not just in simulation.
• High agency — you don't wait to be told what to do.
• Fluent in English.

Why Join Us

Direct collaboration with Stanford and UC Berkeley researchers on VLA models and embodied intelligence. Bay Area location chosen for proximity.