Researcher – Physical AI Evaluation

About the role

MicroAGI is building the data layer for physical AI and needs a researcher to define what good looks like for the next generation of physical AI systems. You will design and run evaluations that are missing from the field, compare model performance on real robots versus simulation, and turn research insights into actionable product improvements.

Key responsibilities

• Lead evaluation and benchmarking research for physical AI models.
• Run existing industry and research benchmarks and create new ones where gaps exist.
• Investigate model behavior on real hardware compared to simulated environments and work to close the gap.
• Research intervention, autonomy, and the boundary between autonomous model actions and human oversight.
• Publish findings in appropriate venues and help shape the team’s academic and public reputation.
• Collaborate closely with research engineers to move ideas from paper to deployment.

Required profile

• PhD in robotics, machine learning, or a closely related field, or an equivalent record of published research.
• Strong track record in evaluation, benchmarking, or applied robotics research.
• Demonstrated ability to work with real hardware, not only in simulation.
• High agency and self‑direction.
• Fluent English communication skills.

Required skills

• Robotics
• Machine learning
• Evaluation and benchmarking methodologies
• Experience with real‑world hardware platforms