Research Scientist / Engineer

Your Mission

• Build and ship learning-based control systems that power humanoid robots in the real world
• Train and deploy policies using reinforcement learning, imitation learning, and multimodal foundation-model approaches with embodied perception
• Hack on simulation and real-world setups to close the sim-to-real gap and move faster from experiment to deployment
• Dig into the full robotics stack to chase down latency, bottlenecks, and weird failure modes
• Create intuitive teleoperation and human-in-the-loop tools to accelerate data collection and evaluation
• Work side-by-side with a small team of engineers and scientists who care about moving fast and building things that last

What You Bring To The Table

• A strong drive to get humanoid robots working in the messy real world—not just in papers or demos
• Ability to take full ownership: spotting problems, proposing solutions, and pushing them through to deployment
• Solid experience with machine learning for robotics (RL, IL, VLA, or related), plus comfort jumping between code, data, and hardware
• Production-level Python or C++ skills
• 3/5+ years of hands-on contributions in robotics, ML, or similar—whether from industry projects, research labs, or your own builds
• Care for detail: clean code, reliable experiments, and datasets that don’t crumble under pressure
• Bonus: you’ve tinkered with humanoids before
• We value exceptional builders over perfect resumes. If you don’t meet every criterion but believe you can have an outsized impact, we strongly encourage you to apply