Atlas VLA Research Intern

Boston Dynamics • Waltham, MA

8d • $30 - $70

About The Position

The Atlas VLA Research Intern joins the Atlas VLA Research Team to develop intelligence that enables Atlas to perceive, reason, and act. The team scales Vision-Language-Action (VLA) models using diverse data sources and training approaches so Atlas can perform challenging dexterous tasks. This is a research-focused internship for PhD-level candidates, with projects aimed at advancing humanoid robot capabilities through experiments and deployment on physical hardware.

The intern will lead a research project on VLA/WAM performance, develop and evaluate data and training methods, and deploy algorithms on Atlas, progressing from simulation to hardware. The role also involves collaborating on integration with low-level robot control, extending data pipelines, and writing production-grade Python or C++ code in existing systems.

Requirements

• Actively pursuing a PhD, preferred, or a research-heavy Master’s in Computer Science, Robotics, Machine Learning, or a related field.
• Expertise in training and deploying imitation learning models (VLAs/WAMs) for robotics.
• Strong experimental hygiene and clear knowledge of the state of the art and areas where the research community can improve.

Nice To Haves

• Good grasp of machine learning best practices, architectures, and scaling laws across different domains.
• Experience troubleshooting and deploying algorithms on physical robot platforms, especially mobile and humanoid form factors.
• Experience working with and contributing to large-scale datasets such as EgoScale or Open X-Embodiment, or with specialized data collection approaches such as Universal Manipulation Interface.
• Experience with large-scale cluster training, including Kubernetes, SLURM, or distributed GPU training, and maintaining high-quality codebases.

Responsibilities

• Lead a research project establishing hypotheses and experiments for VLA/WAM performance, extending benchmarking, training, and evaluation across data sources such as teleoperation, UMI-style data, and VQA.
• Prototype and deploy data collection, mixing, and grading schemes; pre-, mid-, and post-training recipes; and algorithms directly on Atlas, moving from simulation to hardware.
• Collaborate across teams to integrate learned policies with low-level robot control.
• Extend data pipelines that ingest alternative data sources to improve robot robustness.
• Write production-grade Python and C++ code that integrates with existing systems, using AI coding agents to accelerate progress while maintaining code quality.

What This Job Offers

• Job Type: Full-time
• Career Level: Intern