Robot Learning Researcher

DeepReach AI; San Francisco (CA); On-site; USD 150,000 - 230,000; Full time

Job summary

DeepReach AI in San Francisco is building the data infrastructure for robot learning. We seek researchers who can train robot manipulation policies, design experiments, and build reproducible training and evaluation frameworks.

Candidates should have advanced degrees or equivalent experience, strong PyTorch, and hands-on policy deployment to real robots. You will work with frontier labs and publish at top venues, owning your work end-to-end.

Qualifications

• MS or PhD in Robotics, Computer Science, Machine Learning, or related field, or equivalent experience.
• Strong PyTorch skills and hands-on experience deploying policies to real robots.
• Familiarity with major open-source imitation learning or VLA frameworks.
• Solid robotics fundamentals: coordinate frames, kinematics, basic control.
• Rigor in evaluation with controlled experiments and statistics.

Responsibilities

• Train robot manipulation policies (imitation learning, diffusion policies, VLAs) on large-scale human demonstration and robot data, and evaluate them in simulation and on real hardware.
• Design systematic experiments to understand data scale, diversity, and composition effects on policy performance.
• Build reproducible training and evaluation frameworks and standardized benchmarks.
• Turn experimental findings into actionable data standards, collaborating with data and engineering teams.
• Track and reproduce state-of-the-art research to keep the team at the frontier.

Skills

PyTorch

Robot policy deployment

Imitation learning / VLA

Robotics fundamentals

Experiment design & statistics

Education

MS/PhD in Robotics/CS/ML or equivalent

Tools

Open-source frameworks

Job description

We are an early-stage physical AI company building the data infrastructure for robot learning. We work closely with frontier AI labs, robotics companies, and leading university research groups.

What You'll Do

• Train robot manipulation policies (imitation learning, diffusion policies, VLAs) on large-scale human demonstration and robot data, and evaluate them in simulation and on real hardware.
• Design systematic experiments to understand how data scale, diversity, and composition affect policy performance.
• Build reproducible training and evaluation frameworks and standardized benchmarks.
• Turn experimental findings into actionable data standards, working with data and engineering teams.
• Track and reproduce state-of-the-art research to keep the team at the frontier.

What We're Looking For

• MS or PhD in Robotics, Computer Science, Machine Learning, or a related field, or equivalent experience.
• Strong PyTorch skills and hands-on experience taking a policy from training to real-robot deployment.
• Familiarity with major open-source imitation learning or VLA frameworks.
• Solid robotics fundamentals: coordinate frames, kinematics, basic control.
• Rigor in evaluation: claims are backed by controlled experiments and statistics.

Nice to Have

• Research on learning robot policies from human video.
• Human-to-robot motion retargeting.
• RL fine-tuning of learned policies.
• Experience with humanoid or dexterous hand platforms.
• Simulation environment building and sim-to-real transfer.
• Publications at CoRL, RSS, ICRA, NeurIPS or similar, or impactful open-source work.

Why Join Us

• Work alongside frontier AI labs and top university research groups on the field's hardest problems.
• Publish at top venues and contribute to open source, with full support.
• Own your work end to end, from data to models to real-world deployment.
• Early team, high ownership, competitive salary and equity.