Staff / Principal Perception Engineer - Physical AI & Autonomous Robotics — San Francisco Bay Area

We're partnering with an ambitious, well-funded Physical AI company that's developing intelligent autonomous systems capable of operating in complex, unpredictable real-world environments.

Their technology brings together cutting-edge computer vision, deep learning and robotics to solve some genuinely difficult engineering challenges. And unlike many AI companies, success isn't measured purely in model benchmarks. It's measured by how reliably their technology performs in the real world.

They're now looking for a Staff / Principal Perception Engineer to take ownership of critical elements of their perception stack as they continue to develop and deploy their technology at scale.

You'll be working at the intersection of machine learning, computer vision and robotics, developing the intelligence that enables autonomous machines to understand and interact with their surroundings.

This is a deeply hands-on position with significant technical ownership.

You'll be:

• Designing and developing advanced perception algorithms for complex, dynamic environments.

• Building deep learning models for object detection, segmentation, tracking and 3D scene understanding.

• Developing robust sensor fusion solutions using cameras, LiDAR and other sensing technologies.

• Solving challenging problems in localization, spatial awareness and real-time perception.

• Optimizing models for deployment on resource-constrained robotic hardware.

• Taking ownership of perception architecture and influencing key technical decisions.

• Working closely with autonomy, controls and hardware engineers to deliver systems that perform reliably outside the lab.

We're interested in engineers who have already solved difficult perception problems and want the freedom to tackle even bigger ones.

Your background:

• Extensive experience developing perception or computer vision systems for robotics or autonomous platforms.

• Strong Python and modern C++ skills.

• Deep learning expertise, particularly with PyTorch.

• Experience with 3D perception, sensor fusion, object tracking, localization or SLAM.

• A track record of taking algorithms from development through to deployment on real hardware.

• The ability to lead technical direction while remaining an excellent hands-on engineer.

• Experience with CUDA, TensorRT, ROS 2 or real-time edge inference would be a bonus.

Why this is interesting?

Real engineering ownership. You'll be trusted to make important architectural decisions and help shape a core part of the company's technology.

Hard problems worth solving. Dynamic environments, imperfect sensor data, real-time constraints and the challenge of getting intelligent machines to operate reliably in the real world.

Technology with tangible impact. Your work won't just improve a benchmark. You'll see it operating on physical autonomous systems.

An opportunity to make your mark. Join a growing engineering team where your expertise will have a direct impact on both the technology and the company's direction.

This isn't a role for someone looking to move away from engineering into management. It's for someone who loves the technical challenges, enjoys taking ownership and wants to build something exceptional.