3D Reconstruction Architect for Spatial AI

Niantic Spatial, Inc.; Greater London; Hybrid; GBP 70,000 - 120,000; Full time

Benefits offered by this job

Annual bonus

Equity

Comprehensive benefits

London office

Job summary

Niantic Spatial, Inc. in London is seeking a 3D Reconstruction Research Scientist/Engineer to bridge academic research and customer-ready deployment.

You will work on 3D reconstruction, neural scene representations, and spatial AI, turning cutting-edge ideas into practical systems used by real customers. You will collaborate with researchers across CV, ML, graphics, and robotics on large-scale spatial understanding, with scope based on experience.

Qualifications

• PhD in Computer Vision, Robotics, ML, CS, Mathematics, Physics, or equivalent.
• Strong 3D vision research record: reconstruction, depth, SfM, SLAM.
• Multiple high‑profile publications (CVPR/ICCV/ECCV/SIGGRAPH/NeurIPS or equivalent).
• Strong Python programming with PyTorch experience.
• Proven ability to translate research into real features or production pipelines.
• Independent researcher capable of driving projects from idea to publication and handoff.

Responsibilities

• Research & publication — conduct original 3D vision and spatial AI research; publish at top venues.
• Algorithm development — design and evaluate 3D reconstruction, neural scene representations, and related models.
• Research-to-production — transition ideas into customer-facing products with real data and constraints.
• Large-scale experimentation — build and run benchmarks and validate new methods.
• Cross-disciplinary collaboration — work with CV, ML, graphics, and robotics teams.
• Rapid prototyping — implement state-of-the-art methods and deliver working prototypes to customers.

Skills

Python programming

PyTorch

Research independence

3D computer vision

Education

PhD in Computer Vision/Robotics/ML/CS/Math/Physics

Tools

COLMAP

Job description

Niantic Spatial, Inc. in London is seeking a 3D Reconstruction Research Scientist/Engineer to bridge academic research and customer-ready deployment.

You will work on 3D reconstruction, neural scene representations, and spatial AI, turning cutting-edge ideas into practical systems used by real customers. You will collaborate with researchers across CV, ML, graphics, and robotics on large-scale spatial understanding, with scope based on experience.