Embedded Computer Vision Engineer

Our client build autonomous navigation for UAVs - systems that actually fly, in real conditions, without GPS as a crutch. The team is small, the problems are hard, and the code ends up on hardware. 

If you've worked on vision systems that had to work in the field - not just in simulation - this might be worth a look.

What we're looking for

• 5+ years in computer vision, with real experience in robotics, UAVs, or autonomous systems.
• Shipped real-time vision algorithms — detection, tracking, segmentation, or navigation — not just research code.
• Hands-on with edge deployment and model optimization for resource-constrained hardware.
• Strong Python and/or C++. Comfortable with PyTorch or TensorFlow, and OpenCV.
• Master's or PhD in CS, Robotics, EE, or related field.
• You can own a problem and communicate clearly

Nice to have

• Hands-on SLAM experience (ORB-SLAM, VINS, or similar).
• Familiarity with MAVLink protocol, ArduPilot, or PX4.
• Real-world UAV testing - not just Gazebo.
• Publications or open-source contributions in vision or robotics.

What you'll do

• Design and optimize CV algorithms for localization, mapping, and navigation -own this end-to-end.
• Work on detection, segmentation, SLAM, 3D scene reconstruction, visual odometry, and sensor fusion (cameras, LiDAR, IMUs).
• Deploy and optimize vision models on embedded platforms — Jetson, Raspberry Pi class devices.
• Integrate CV modules with flight controllers and navigation systems via MAVLink/ArduPilot/PX4.

• Work closely with hardware and software teams - this is a cross-functional role, not a silo.

  Run experiments, analyze results, iterate. Ship things that fly.