CareerPlan

Sr. Computer Vision Engineer (3D Semantic Scene Understanding)

MunichFull-time2026-03-29 → 2026-11-16

Core

Lead the development of the spatial reasoning engine for an agentic AI platform, focusing on 3D Semantic Reconstruction, Geometric Deep Learning, and Agentic Inference to transform unstructured data into Spatial Knowledge Graphs.

Role type

Senior IC machine-learning engineer (3D semantic scene understanding)

Builds

Spatial Knowledge Graphs, 3D scene representations, and agentic inference workflows for the AEC industry

Domain

Architecture, Engineering, and Construction (AEC) + 3D Computer Vision

Deliverable

production ML models

Required skills

3D Computer Vision, Geometric Deep Learning, 3D Reconstruction, Multi-modal Fusion, Knowledge Graph Augmentation, Agentic Inference, Physics Engine Integration, Software Design

Preferred skills

Agentic AI implementations (GraphRAG, Langgraph/LlamaIndex), Physics engines (NVIDIA Isaac Gym, MuJoCo, PyBullet), Startup environment experience

Technologies

Open3D, PyTorch 3D, TSDF, Multi-view stereo, Mesh-fitting

Responsibilities

Develop algorithms for 3D scene representation prioritizing geometric primitives and semantic labels; Architect systems to fuse panoptic segmentation with 3D models; Automate augmentation of 3D spatial data to Spatio-Temporal Knowledge Graphs; Design agentic workflows for complex reasoning tasks on STKG; Implement methods to identify affordances within 3D volumes; Deploy SOTA models and representations into production use-cases

Seniority

Senior, hands-on IC