Develop and maintain software at the intersection of machine learning and robotics, with a focus on robot grasping — determining which gripper configuration and tools to use.

Integrate, improve, and harness SOTA ML models for our task domain. Modify the architectures, create training pipelines, deploy to robots, and continually iterate.

Collaborate closely with a small, highly technical team (currently two engineers).

Contribute to simulation-based development and testing of grasping and manipulation systems.

Participate in code reviews, technical discussions, and cross-functional problem-solving with perception and motion planning teammates.

Extensively leverage AI coding tools and modern development workflows to maximize individual and team output.

Hands-on experience with machine learning — you have implemented complete ML systems before, understand the terminology, and are comfortable working with medium-scale cloud-based training pipelines.

Experience applying ML in the context of a robotic system is strongly preferred; a background that spans both the perception and motion planning domains is ideal.

Strong Python skills are essential. Cloud knowledge (especially GCP) is preferred.

Familiarity with classical robotics concepts (motion planning, geometry, perception)

Experience with simulation tooling (Isaac Sim, MuJuCo, etc) for robotics development is a meaningful advantage.

BS required; MS preferred in Robotics, Computer Science, or a related field.

Active use of AI coding assistants and modern development tooling in your day-to-day workflow is preferred.

Ability to work cross-functionally and communicate clearly across engineering teams with different specializations.

Willing to work in the office from our Charlestown, MA location at least three days per week.