About Lumen Labs

Lumen Labs is building the cognitive layer for physical AI. Nearly every major robotics effort today teaches machines by showing them thousands of hours of human teleoperation. We think that paradigm has a ceiling, and that the path to robots that genuinely adapt runs through a different kind of learning and we are trying to figure out what that is. We are primarily focused on nature inspired AI architectures for unstructured environments (construction sites, mining, oil pipelines, battlefield — places where no pre-training will help).

The founders are Harry (2x founder, Thiel Fellow, previous 1517 Fund) and Daniel (OSV Fellow, previously Wolfram Institute, The Residency) with a background in cognitive science, robotics and sensors. Pre-seed from top VC funds and operators at robotics/AI companies (Anthropic, Waymo, etc.).

Location: In-person in San Francisco. We're happy to figure out visa and relocation for the right candidate. 🇺🇸

About the Role

You'll own the hardware and infrastructure that keeps our research platforms running. Today that's wheeled navigation, we're dipping our toes into drones, and we're eyeing weirder platforms like pipeline inspection. This means maintaining and extending ROS nodes across onboard and offboard compute, integrating new sensors (audio, speech, vision), managing the safety pipeline during autonomous runs, and keeping the data logging pipeline clean and reliable. As we scale our physical prototypes, you'll architect the ROS infrastructure.

You’ll need to work closely with our model engineers to figure out how to bridge the sim2real gap for specific robots and tasks. You won't need to set up the sims yourself, but you'll be the one getting trained policies onto real hardware. You will need to work closely with the rest of the team to figure out where the robot and environment differs from the sim and how to mitigate those differences.

We provide competitive compensation (salary + equity). We also have health insurance and provide visa/relocation support.

What You'll Do

• Build and maintain the ROS2 workspace for robot control: the node architecture, topic graph, and inter-process communication across robot and laptop compute
• Deploy trained policies to the robot: observation assembly, inference at a fixed control rate, command publication, logging
• Integrate and calibrate sensors: LiDAR, depth camera, IMU, encoders, microphone, and future additions. Calibrate the sensors against physical measurements.
• Diagnose and localize robotics problems and figure out the appropriate firmware/hardware fix.
• Own the safety pipeline and real-world safety behavior during autonomous experimental runs
• Help the team bridge the gap between sim2real: characteristic testing on hardware (system identification), domain randomization, creating digital twins, fine-tuning on hardware
• Manage hardware bring-up, Docker environments, and on-device dependencies.
• Run experimental trials, maintain logging infrastructure, and ensure data quality

<aside> 🔥

You'll thrive here if you've built and run real robots of more than one kind, and have done off-the-beaten-path side projects along the way.

</aside>

Requirements

• Comfortable with ROS2 (Humble) across different robot platforms