Our client is building humanoid robots and learning systems for the physical world. Their focus is simple: make light work. They’re looking for someone who lives at the intersection of simulation and real robots — someone who can close the sim-to-real gap on contact-rich manipulation and help train policies that actually work on hardware.
What you’ll do:- Build high-quality simulated tasks, assets, and evaluation environments
- Train and evaluate manipulation policies
- Develop robust methods for transferring learned behavior from sim to physical robots
- Create tools that accelerate data generation, training, and evaluation
- What they’re looking for 4+ years in robot simulation or robot learning
- Experience deploying learned behaviors on real robots
- Strong fundamentals in simulation, system identification, and sim-to-real
- Proficiency in Python and a modern ML framework
Bonus points for experience with deformable manipulation, simulation infrastructure, or large-scale training systems.
This is an early-stage team in New York City building the next generation of humanoid capability. If you want to ship systems that move from simulation into the real world, reach out.
