Help us build the missing layer of robotic dexterity.
Work with us to turn skilled robotic manipulation into a real-world capability.
We are a full stack robotics company, and are looking for candidates with expertise at all levels of the robotic manipulation stack. This includes, but is not limited to, expertise in the following areas of robotics, and their many crossovers:
Learning
- Reinforcement Learning (DeepRL, sim2real, Domain Randomization, RL on real hardware, Offline RL, RL from Human Feedback, etc.)
- Imitation Learning / Behavioral Cloning (Diffusion Policies, Flow Matching, Action Chunking, etc.)
- Classical learning methods (supervised / unsupervised learning) applied to robotics problems
- Interaction between various learning paradigms
Control
- Classical feedback control (position, velocity, impedance, etc.)
- Optimal control (MPC and similar methods such as MPPI, iLQR, etc.)
- Interaction between controllers and learned policies (RL, BC, etc.)
Manipulation
- Kinematic chain forward, inverse and differential kinematics, kinematic chain dynamics
- Path and trajectory planning, collision avoidance
- Fundamentals of 2D and 3D computer vision
- Multimodal robotic manipulation combining vision, tactile and proprioceptive sensing
Design and Mechatronics
- CAD for compact, serviceable kinematic assemblies
- Drivetrain design: actuators, transmission mechanisms, etc.
- Prototyping and manufacturing methods (machining, 3D printing, etc.)
- Microcontroller programming and firmware
Integration and Operations
- Integration of real robotic hardware, simulation and visualization
- Virtual environments, dockerization and containerization
- Version control and continuous integration
We invite exceptional candidates at all levels of education (BS, MS, PhD) and experience to get in touch with us. The ideal candidate will appreciate the energy and do-it-all attitude of an early stage startup, and is expected to leave a profound mark on how the future of the company is shaped. All positions are based in New York City.
