Sieve
Member of Technical Staff, Computer Vision (Perception)
San Francisco · On-site · Staff · Full-time · Posted Oct 26, 2026
- Experience
- 3+ years building production perception or computer vision systems for robotics, AR/VR devices, autonomous vehicles, or similar.
About the job
Build production perception systems that transform egocentric, robot, and sensor footage into high-quality training data for robotics and world-model teams.
What you'll do
- Build perception pipelines for depth, pose, tracking, segmentation, and 3D reconstruction, using techniques such as SLAM and visual odometry.
- Fine-tune vision and vision-language models, and develop data QA systems, benchmarks, and evaluation frameworks.
- Work with AI labs and robotics teams to translate data needs into deployed pipelines and iterate based on results.
What they ask for
- 3+ years building production perception or computer vision systems for robotics, AR/VR devices, autonomous vehicles, or similar.
- Hands-on experience with SLAM, visual-inertial odometry, multi-view geometry, or 3D reconstruction.
- Strong Python skills and fluency with PyTorch.
- Experience training, fine-tuning, or deploying modern deep learning models for vision.
Nice to have
- C++ experience.
- Experience with NeRFs or Gaussian splatting, egocentric or teleoperation data, or relevant publications or open-source work.
Tools
Python, PyTorch, C++, SLAM, visual-inertial odometry, NeRFs, Gaussian splatting
AI focus
Fine-tune vision and vision-language models for embodied tasks, and create training data, benchmarks, and evaluation systems for robotics and world models.
Model-written summary · the original posting is authoritative