Atlas ML Software Engineer - Large Scale Behavior Cloning

Boston Dynamics is looking for an ML Software Engineer to build the tools and infrastructure powering large-scale behavior cloning for the Atlas humanoid robot. The role spans data pipelines, training infrastructure, evaluation systems, and developer tooling to accelerate the collect-train-evaluate loop. You will collaborate with research scientists and engineers to improve policy performance and data quality across both hardware and simulation.

• Collaborate with research scientists and infrastructure engineers on R&D initiatives.
• Develop tools to query and gain insights into data, training, and testing performance metrics.
• Analyze and track failures in experimental robot runs.
• Work with robot simulation software to create scenarios, tests, and data collection tools.
• Build internal tools and platforms to accelerate the collect, train, and evaluate loop.
• Provide visibility into policy performance, data quality, and collection gaps.
• Support evaluation and rapid experimentation across hardware and simulation.
• Maintain and improve code repository, including CI/CD, testing, and deployment.

• 3+ years of experience developing software for robotics, ML infrastructure, or related fields.
• Strong proficiency in Python; familiarity with C++.
• Experience building internal tools, dashboards, or GUI applications.