Lead Engineer

Every sensor. One clock.

AnchorCam Stereo is a head-worn camera that records the first-person data robots learn from. Two 1920×1200 global-shutter cameras sit 70 mm apart. One hardware trigger fires both cameras and the IMU. Recordings go to microSD, and the camera keeps working without a laptop.

The timing works on the boards we have built. The next job is making every hour of recording something a robot team can train on without cleaning it first. That is the work this role leads.

Backed By #

Kay Kyungsik Woo · Founder & CEO of MVL Labs , the company behind TADA, a mobility platform on MVL Chain (2M+ users, 200K+ drivers, $40M+ raised, ~250 employees globally).

What You’ll Build #

• Stereo calibration that holds up across units, remounts and a day on someone’s head, not only on a checkerboard at a desk
• Stereo + IMU tracking as a quality check on every recording, with honest reporting of where it fails
• The pipeline from microSD or Wi-Fi to a dataset a customer can load
• Measurement tools that show how far each recording is from what we promise, before a customer finds out
• ISP tuning on the Rockchip RV1106 with real sensors, where both cameras have to agree
• Putting the on-board NPU to work within its power budget

Requirements #

• You have shipped image processing that ran on a device, not only in a notebook
• Comfortable at the sensor end of the stack: raw frames, exposure, timing, calibration
• Can go full-time starting now

Nice to Have #

• Multi-camera calibration or visual-inertial tracking in something that left the lab
• MIPI CSI-2, sensor bring-up or ISP tuning
• Experience with robot-learning datasets or teleoperation data

What We Offer #

• 3% to 4% equity · the first engineering hire, and the grant is sized to match
• Hybrid-remote · work remotely, but able to be in Seoul periodically for in-person collaboration
• Hardware that already exists. Boards are built and recording, so your first week goes on real data

Interested? → hello@mutual.solutions