Computer Vision Engineer
Overview:
In this R&D role at Tutor, you will own perception for robot manipulation, guiding how vision systems enable reliable real-world operation. You'll work with cross-functional teams to define success criteria, build and integrate detection and 3D understanding, and maintain datasets and evaluations on physical robots. Your work sits at the intersection of camera geometry and learned models to improve robotic perception. This role offers hands-on impact from research to deployment, shaping how perception drives manipulation. Compensation / Benefitsfully employer-covered health insurancemanaged 401(k)regular in-office mealssocial eventscollaborative, low-ego cultureownership and problem-solving orientation
Responsibilities:
- Define approach and success criteria for ambiguous perception problems in robot manipulation
- Build and integrate capabilities: detection, segmentation, tracking, pose estimation, 3D scene understanding
- Own camera calibration, coordinate transforms, synchronization checks, and degradation monitoring
- Create reproducible datasets and evaluations across object changes, lighting, viewpoint, occlusion
- Select and adapt geometric methods and learned models based on accuracy, latency, robustness
- Deploy and maintain perception in the robot stack, diagnose failures from data and experiments
- Improve evaluation tools and documentation, participate in design reviews, mentor colleagues
Key requirements:
- Evidence of independently owning a substantial computer-vision system from problem to tested integration
- Strong Python engineering with maintainable code, tests, and reproducible experiments
- Practical command of camera models, calibration, coordinate frames, and 3D geometry
- Experience with geometric vision or learned visual models and knowing when to use each
- Ability to build representative evaluations, isolate failure modes, distinguish model vs data artifacts
- Ability to communicate technical tradeoffs and coordinate with engineers from other disciplinescommunication of technical tradeoffscross-disciplinary collaborationmentorship
OpenCVOpen3DPCL