Software Engineer, Computer Vision

This is a fully remote position, open to applicants in United Kingdom.

📋 Description

• Enhance automated optical inspection systems to identify missing, rotated, inaccurately positioned components, and poor soldering.

• Develop and implement models to locate components in bins and trays, center them on the nozzle, and recognize clumped or dropped picks.

• Execute fiducial detection, calibration, and image alignment to ensure precise placements.

• Determine the data to be collected, create labeling tools, and transform operator feedback into training data.

• Collaborate with high-speed cameras and the GPU capture pipeline to bring vision processing closer to real-time.

• Monitor production accuracy, pinpoint sources of error, and verify modifications against historical data.

• Partner with AI agents and assist in expanding CircuitHub's factory-scale robotics platform, The Grid.

⛳️ Requirements

• You have successfully launched computer vision applications into production, extending beyond a notebook environment.

• Proficient in software engineering with expertise in Python, C++, or a similar programming language.

• A blend of classical vision techniques (geometry, calibration, template matching) and deep learning; you understand when to apply each approach.

• Familiarity with industrial or machine-vision cameras is advantageous.

• You utilize coding agents to enhance efficiency and take full responsibility for their outputs.

• You have an appreciation for the physical aspects of the environment: lighting, lenses, and components rotated by 90°.

• Experience with OpenCV, PyTorch and Lightning, ONNX Runtime on GPU, and libtorch from Haskell.

• Knowledge of Basler GigE cameras through pylon and high-speed cameras with GPUDirect capture.

• Proficient in Python, Haskell, PostgreSQL, Nix, and NixOS.

🏝️ Benefits

• Equity

• Competitive salary