Design the data backbone for next-generation robotics and Physical AI systems. In this role, you'll translate complex robotic behaviors into structured data requirements, enabling breakthroughs in perception, action modeling, and world understanding. You’ll work at the intersection of robotics, machine learning, and data engineering to build pipelines that drive training and evaluation for autonomous systems.
Key Responsibilities
- Define how real-world robotic tasks translate into data needs for training policies, perception models, and environment simulation.
- Build and test prototypes for vision-language models, world models, and action representations to identify critical data dependencies and quality metrics.
- Leverage simulation platforms and digital twins to generate and validate datasets that reflect real sensor inputs, physical constraints, and operational scenarios.
- Partner directly with robotics and ML teams at client organizations to establish data specifications, collection methods—including egocentric capture and imitation learning—and performance benchmarks tied to model outcomes.
- Lead technical scoping and pilot engagements, designing experiments that demonstrate measurable improvements from data pipelines and annotations.
- Collaborate with internal engineering and data teams to develop scalable annotation workflows, data QA processes, and reusable pipeline architectures.
- Create reference designs, playbooks, and demonstrations for common use cases such as manipulation, navigation, teleoperation, and human-robot interaction.
- Inform the evolution of a dedicated Robotics & Physical AI data platform by feeding real-world insights into product and tooling priorities.
- Represent the organization at industry forums, sharing best practices in robotics data strategy and helping grow a broader technical ecosystem.
What You Bring
- Proven experience in robotics, computer vision, or embodied AI, with hands-on work in simulation or real robotic systems.
- Ability to connect physical tasks to model behavior, data structure, and evaluation criteria—communicating trade-offs clearly across technical and non-technical stakeholders.
- Familiarity with imitation learning, teleoperation, world models, or synthetic data generation in robotics contexts.
- Minimum of three years of development experience in Python and a systems programming language such as C++ or Java.
- Comfort navigating undefined problems, rapidly prototyping minimum viable solutions, and iterating with customers.
- Strong technical communication skills—able to lead whiteboarding sessions, facilitate workshops, and align with foundation model and robotics teams on data design.
- Project leadership experience: driving initiatives from concept to delivery while coordinating across teams and balancing technical with business goals.
Preferred Background
- Hands-on work with humanoid robots, manipulators, or mobile robots, or advanced simulation environments like Isaac or Omniverse.
- Experience designing large-scale datasets, annotation schemas (e.g., affordances, action labels), or evaluation frameworks for robotics models.
- Previous technical consulting, pre-sales engineering, or solutions architecture in robotics or autonomous systems.
- Contributions to open-source robotics or ML projects, technical publications, or public demos showcasing data or system design.
Environment & Impact
This position offers full remote work within the United States, placing you at the forefront of robotics innovation. You’ll collaborate with leading technology teams, influence the direction of data infrastructure for Physical AI, and contribute to building a secure, scalable platform used across the robotics ecosystem. Your work will directly shape how machines learn from data in real-world contexts.
