WHAT YOU WILL BE DOING
Develop and optimize visual inspection algorithms for defect detection, anomaly detection, classification, and quality validation using deep learning
Optimize model inference for GPU deployment, leveraging CUDA, TensorRT, and related acceleration frameworks
Collaborate with Mechanical engineers to design illumination setups that maximize inspection accuracy and robustness
Build and maintain data pipelines for model training, evaluation, and continuous improvement
Partner with platform team to establish MLOps practices for model versioning, experiment tracking, automated retraining, and production model monitoring
Harden inspection solutions for production reliability, including monitoring, alerting, and graceful degradation
Work with service engineering and field teams to deploy inspection solutions and support customer rollouts
- Define metrics and benchmarks to measure inspection accuracy, throughput, and reliability
- WHAT YOU WILL BRING
- MS or PhD in Computer Science, Electrical Engineering, or a related field, or the equivalent in experience with evidence of exceptional ability.
5+ years of relevant experience in computer vision and/or machine learning
Strong programming skills in Python
Deep experience with PyTorch for model development and training
Experience optimizing ML models for GPU inference in production environments
Track record of shipping ML/CV models from prototype to production
- Experience with image acquisition, camera systems, and sensor integration
- IT WOULD BE GREAT IF YOU HAD
- Knowledge of lighting and optics for machine vision (diffuse/directional illumination, lens and filters)
- Experience with industrial camera systems and standards (GigE Vision, GenICam, CoaXPress)
- C/C++ experience for performance-critical components
Experience with MLOps tooling (MLflow, Weights & Biases, Kubeflow, or similar)
Experience with data annotation, labeling workflows, and active learning strategies
Experience with ROS2
Understanding of manufacturing processes and quality control methodologies
Publications or patents in computer vision, deep learning, or related fields