3D Vision Engineer – Spatial Intelligence

We are looking for a 3D Vision Engineer – Spatial Intelligence to advance the geometric foundations of our spatial intelligence stack across edge devices and cloud. Our robots have to know where they are in environments that don’t sit still: shelves get restocked, layouts change, lighting shifts, objects move. You’ll be responsible for the robustness of the 3D vision pipeline that holds up under all of it (VIO, SfM, volumetric reconstruction, and large-scale visual relocalization), running on edge devices and in the cloud. Re-localization in changing environments is the central problem. Your work decides how reliably the fleet stays oriented as the world around it evolves.

What the role covers

• Ship VIO, SLAM, SfM, and visual relocalization that survive real-world environmental change, on resource-constrained edge devices and at cloud scale.
• Architect volumetric reconstruction and rendering (Gaussian splatting and beyond) across diverse input modalities. This is a core lever for robust relocalization.
• Combine classical 3D vision with data curation and simulation to build large-scale datasets for 3D foundation models.
• Drive the integration of geometric vision with learned perception for robust spatial reasoning in dynamic environments.

What it takes

• MS or PhD in Computer Vision, Robotics, or related field, with strong multi-view geometry foundations.
• Production experience with at least one of the following, deployed on real robots or fleets: VIO/SLAM, SfM, or large-scale visual localization.
• Multi-sensor calibration and nonlinear optimization (Ceres, GTSAM, or g2o).
• Volumetric reconstruction or neural rendering experience (Gaussian splatting, NeRF variants) is a plus.
• Familiarity with state-of-the-art research and an instinct for what’s worth porting from a paper into production.