Staff Robotics Software Engineer

Staff Robotics Software Engineer (Physical AI / Embodied Intelligence)

The Opportunity: Scaling Production-First Physical AI

While much of the robotics industry is focused on research and lab prototypes, our client is a production-first leader in Embodied Intelligence. They have 50+ autonomous robotic cells already live in heavy manufacturing environments, solving complex physics problems that traditional simulators struggle to capture.

​

We are looking for a Staff Robotics Software Engineer to provide high-level technical leadership. This is a hands-on IC role where you will architect the bridge between a production C++ robotics stack and a proprietary neural world model.

​

The Role

You will define the technical direction across the entire robotics software stack, perception, planning, controls, and execution. Your focus will be integrating modern AI techniques (Reinforcement Learning, learned planners, and foundation models) into a classical robotics framework to enable autonomous systems to perform increasingly complex real-world tasks.

​

* Architect & Build: Lead the development of core robotics software, from hardware abstraction layers to high-level motion planning and trajectory generation.

* Technical Leadership: Guide a team of 8 engineers, driving engineering excellence and cross-functional strategy without the overhead of people management.

* Production Excellence: This isn’t a lab role. You will be writing and shipping production code that operates reliably on 5,000lb industrial machines in high-stakes environments.

* Bridge the Gap: Evaluate how to best use neural world models to complement classical robotics approaches for tasks like adaptive sensor-driven motion and multi-robot coordination.

​

* Experience: 8+ years of experience in robotics software development. You have a track record of architecting and delivering large-scale systems from concept to deployment.

* C++ Expert: Deep expertise in Modern C++ (14/17/20) is non-negotiable; Python proficiency is strongly preferred.