Senior Staff ML Engineer- Atlas Agents Team

Boston Dynamics•Waltham, MA

16d

About The Position

The Senior Staff ML Engineer will work on the Atlas Applications team to build and integrate systems that enable Atlas humanoid robots to perform complex tasks in dynamic environments. The role spans the agentic system powering Atlas, including harness design, context engineering, and agent benchmarking. The engineer will partner with ML engineers, product teams, interaction designers, behavior and controls experts, and customers to develop Atlas' planning and reasoning system and integrate it with appropriate tools, data, and prompts. The position also contributes to end-to-end capabilities and a strong engineering culture through reviews, testing, documentation, and thoughtful system design. Candidates should bring a strong full-stack software engineering background, spanning backend development and LLM integrations, and be comfortable owning problems in fast-moving, ambiguous environments. The position is open to qualified candidates eligible to work in the United States; visa sponsorship is not available. The stated annual base pay range is $180,000 to $220,000, with pay depending on individualized factors including internal equity, job-related knowledge, skills, and experience.

Requirements

At least 5 years of professional software engineering experience, with a strong full-stack background spanning backend development and LLM integrations.
Demonstrated ability to design and integrate APIs and third-party services in complex distributed system environments.
Strong fundamentals in data structures, algorithms, distributed systems, and software architecture.
Proficiency in modern backend technologies such as TypeScript or Python and data stores such as Postgres or Redis.
Demonstrated ability to ship reliable, maintainable, and well-tested software while using agentic coding tools such as Claude Code.
Comfort working in fast-moving, ambiguous environments and owning problems end to end.
Must be eligible to work in the United States; visa sponsorship is not available.

Nice To Haves

Experience with VLM-backed agentic systems, including context engineering, tool use, and model selection or fine-tuning; for example, building a coding agent.
Experience building model and harness benchmarks, including gym environments, datasets, tooling, or leaderboards for VLMs and/or agents.
Experience working with robotics systems, computer vision models, and real-time data pipelines.
Proficiency in modern frontend frameworks such as React.
Familiarity with observability platforms such as OpenTelemetry.

Responsibilities

Work on Atlas' agentic capabilities using VLMs, tool calls, context engineering, and memory to support real-world applications.
Design and implement observability, logging, and tracing infrastructure to provide visibility into system behavior and performance.
Develop and use evaluations and benchmarks to measure reasoning quality, safety, reliability, and generalization across tasks and environments.
Develop operator dashboards, task configuration UIs, and diagnostic interfaces to help users build and deploy Atlas applications.
Collaborate with behavior, controls, and product experts to ship end-to-end capabilities.
Engage with customers and internal stakeholders to understand real-world use cases and ensure solutions are practical, reliable, and impactful.
Contribute to engineering culture through code reviews, testing, documentation, and thoughtful system design.