Zivid manufactures hardware and software for 3D machine vision in autonomous industrial cells. The tech stack—Python, C++, CUDA, PyTorch, TensorFlow, OpenCV, Open3D—reflects a compute-heavy vision company building real-time 3D reconstruction and ML model inference at the edge. Active projects span algorithm development, compute optimization, and CI/CD scaling; pain points cluster around test flakiness and CI queue times, indicating infrastructure maturity challenges as they ship new products and scale manufacturing.
Zivid is a Norwegian hardware and software company founded in 2015 that designs structured light-based 3D color cameras for industrial automation. Their product line includes the Zivid 3, 2+ R-series, and 2 cameras, paired with an SDK and Studio GUI for integration into robot cells, cobots, and autonomous picking systems. Primary applications are random bin picking, piece picking, assembly, and machine tending. The company operates across engineering, manufacturing, and product teams, with hiring currently concentrated in Norway and focused on mid- to senior-level engineering roles.
Core languages: Python, C++. GPU compute: CUDA, OpenCL, Halide. ML frameworks: PyTorch, TensorFlow. Vision libraries: OpenCV, Open3D. Infrastructure: Docker, Kubernetes, Terraform, Ansible. CI/CD: GitHub Actions, Jenkins, GitLab CI/CD.
3D reconstruction algorithms, ML model integration, compute pipeline optimization, CI/CD redesign, embedded system testing, hardware-in-the-loop validation, and new product introductions.
Other companies in the same industry, closest in size