Robotics Research Infrastructure
tutorintelligence · Watertown, MA · onsite
USD 100,000 to 300,000 · period not stated
Responsibilities:
- Own the technical approach and ongoing reliability of a major system supporting robot data collection and experimentation.
- Build and maintain robot-side and backend services, data pipelines, and tools that reduce operator effort and improve experiment throughput.
- Make system health, data completeness, sensor timing, and experiment progress observable; turn failures into actionable diagnostics.
- Improve deployment, configuration, testing, recovery, and rollback so changes and operations are repeatable.
- Diagnose incidents across device interfaces, Linux, networking, concurrency, and storage; deliver durable fixes and verify their effect.
- Prioritize work within team goals using evidence from researchers and operators, and communicate scope, risks, and tradeoffs.
- Improve engineering tools and runbooks, contribute useful technical reviews, and help colleagues operate and extend the system.
Requirements:
- Evidence of independently owning a substantial software system through design, deployment, operation, and difficult debugging.
- Strong Python software engineering, including maintainable code, automated tests, concurrency, and service interfaces.
- Practical Linux and networking knowledge, and experience deploying and operating services.
- Understanding of distributed-system failure modes, data integrity, and safe recovery from partial failures.
- Experience building observability and using it to diagnose incidents and improve reliability.
- Ability to work across unfamiliar system boundaries, define an approach to ambiguous problems, and collaborate with people who use the system daily.
Nice to have:
- Robotics, device fleets, teleoperation, edge computing, or synchronized sensor pipelines.
- C++ or Rust, containers, cloud infrastructure, infrastructure as code, or ML deployment.
- Operator tools and experiment platforms; effective use and critical review of AI-assisted engineering.
- Establish useful health and data-quality checks, operating documentation, and a prioritized plan for the agreed system.
- Resolve a recurring failure with a measured improvement in reliability or useful data collection.
- Make deployment and recovery repeatable and enable researchers and operators to diagnose routine issues