We are looking for a Senior Autonomy Software Engineer with 5+ years of experience to own the autonomy and perception stack for our autonomous underwater vehicle platform from the ground up. You'll join a ~7-person team building attritable UUVs that operate in radio-denied environments, making all navigation, detection, and targeting decisions independently. This is a greenfield opportunity: you'll architect and ship production autonomy software on hardware (Nvidia Jetson Orin AGX) that has never been deployed on an autonomous underwater vehicle before. We need someone who thrives in early-stage, high-urgency environments and has actually built and deployed autonomy systems on real-world platforms. You'll report directly to the Head of Autonomy and Software and help shape the technical direction of the entire autonomy org.
What will you be doing?
Designing and implementing the full autonomy stack from scratch: mission planning, path planning, object detection/recognition, obstacle avoidance, and autonomous decision-making for underwater operations.
Developing both ML-based and classical perception pipelines that run efficiently on power-constrained embedded compute (Nvidia Jetson Orin AGX).
Taking the vehicle from prototype to executing real 3-day autonomous missions within your first 90 days.
Supporting field testing: deploying scale models in dive pools, running sea trials in the ocean, and performing live telemetry debugging from shore.
Owning your system end-to-end as a Responsible Engineer: from requirements and architecture through simulation, field validation, and production deployment.
- 5-15 years of experience in autonomy/perception/robotics software engineering, using C++ and Python
- Shipped autonomy/perception software on physical autonomous vehicles, UAVs, or robots
- Built autonomy system from scratch (0-to-1), not just inherited mature codebase
- Worked at a fast-paced startup (e.g., Waymo, Aurora, Shield AI, Anduril, Chaos Industries)
- Top-tier school (e.g., MIT, Berkeley, CMU, Stanford)
- ML/deep learning for perception (object detection, tracking, classification)
- Path planning, sensor fusion, state estimation, or classical CV
- Experience with compute-constrained platforms (e.g., Nvidia Jetson)
- BS+ in CS, Robotics, EE, or related technical field
- Hands-on field data debugging and end-to-end implementation experience