Medtronic is looking to build and scale image analytics MLOps capabilities to enable visual inspection, automation, and insights from visual data across its diverse product portfolio in manufacturing.
Requirements
- Proven expertise in AWS cloud services, especially SageMaker, S3, EKS, Lambda, and CloudFormation/Terraform
- Hands-on experience with Kubernetes and containerization (Docker, Helm, GPU orchestration) for AI/ML model deployment
- Experience developing end-to-end ML pipelines, including CI/CD, experiment tracking, and model/data versioning (e.g., MLflow, Sagemaker)
- Strong background in Python and sufficient understanding of AI/DL frameworks (PyTorch, TensorFlow)
- Demonstrated ability to deploy computer vision models in production, including optimization for GPU inference (TensorRT, ONNX Runtime, Triton)
- Knowledge of model monitoring and observability for drift, performance, and compliance in regulated environments
- Experience with edge deployment (e.g., NVIDIA Jetson/Orin, DeepStream, or hybrid cloud/on-prem architectures)
Responsibilities
- Design, implement, and maintain model training, evaluation, and deployment pipelines for computer vision applications, leveraging AWS SageMaker, Triton Inference Server, and Kubernetes-based orchestration.
- Partner with Vision, Manufacturing, and Automation teams to translate feasible models into production-ready inspection systems.
- Build scalable image data pipelines for ingestion, annotation, and curation, ensuring datasets are reliable, versioned, and reproducible.
- Integrate ML services into manufacturing workflows, bridging OT (Operational Technology) with IT/Cloud platforms.
- Implement model observability and monitoring solutions, tracking performance, drift, latency, and GPU utilization for continuous improvement.
- Collaborate on continuous improvement initiatives, including CI/CD for ML models, reproducibility, infrastructure-as-code, and workflow automation.
- Engage with external partners and internal platform teams to align on standardized architecture for analytics for manufacturing imaging
Other
- Minimum of 4 days a week onsite
- Strong collaboration skills across Data Science, Manufacturing Engineering, and IT functions
- The ability to problem solve and troubleshoot machine learning systems
- Familiarity with manufacturing automation systems (PLC, SCADA, OPC-UA, MQTT) and integrating ML services into production lines
- Familiarity with Industry 4.0 concepts, including IoT integration and digital twins for manufacturing
