Robotics Engineer

Humans in Motion (HIM) is a Sports Robotics Research Lab in San Francisco. We are building Adam, the world's best athlete. He will win an Olympic event at LA 2028.

We believe major technological breakthroughs are best shown through sports. Chess with DeepBlue. Go with AlphaGo. Sports in the physical world and Adam.

Adam is 3'11". He's learning every sport on Earth. He already dances, throws spirals, boxes, and kicks. He also knows how to take pictures with a Polaroid camera. Every week we ship a new skill, take him out into the world, and put him alongside the best athletes alive so that he can learn from the best.

Think of us as Mercedes AMG in F1, except the driver is a robot, and the track is every playing field on the planet. Our mission is to win at sports, educate the world about robotics, and eventually share an Adam with every household.

We are self-funded for maximum speed and creativity and just getting started. See what a full skill loop looks like. See our weekly changelog.

You will take sports from concept to reality. That means: pick a skill, build or fine-tune the model, train it in sim, deploy it on real hardware, and then take Adam to a sporting event or collaborate with a world-class athlete to demonstrate it. That's one loop. Then you do it again next week.

We ship a new sport based on what's happening in the world. The Super Bowl? Adam learns to throw a spiral. The NBA playoffs? He's shooting jumpers. You will co-design this roadmap with us. If there are specific athletes you'd like to collaborate with — even better.

This is a great challenge because much of the infrastructure is being built first-hand. You'll be working at the frontier of sports robotics, a field we're pioneering together. We open-source our work so others can learn and grow the ecosystem.

• Own the full skill loop: conception, simulation, training, real-world deployment, athlete collaboration
• Build and fine-tune ML models for dynamic athletic movement — training custom models or riding the best foundation models (e.g. pi) and adapting them for sports
• Work across the sim-to-real pipeline: video analysis, motion retargeting, IsaacLab/Mujoco training, reward function design, deployment on hardware
• Run concurrent training experiments on cloud GPUs, iterate fast, ship weekly
• Collaborate with the world's leading athletes across sports — attend sporting events to demonstrate what's possible
• Buy and integrate best-in-class hardware from partners like Booster and Unitree. Unlimited budget for best-in-class tooling. No institutional red tape.

• Full-stack roboticist. You've built robotics software end to end — from model to metal.
• ML-forward. You want to build the best models for sports, or know how to ride the best existing models and fine-tune them for novel athletic tasks.
• You might be a mechanical or electrical engineer who has spent time in IsaacLab, or a researcher shipping learned policies in the real world. Either way, you ship.
• Love for winning, sports, respect and appreciation for the greats. You look forward to collaborating with professional athletes.
• Interest in attending live sporting events — not just to watch, but to show the world what Adam can do.
• Experience with sim-to-real, reinforcement learning, imitation learning, or diffusion-based policies is a strong plus.

Email Adam at adam@himrobotics.com or DM @himrobotics on X or Instagram. Show us what you've built and why you are exceptional. We look forward to meeting you.