MLOps Engineer
Contract
London based (hybrid or onsite depending on preference)
Outside IR35
£600 to £650 per day

I am supporting a client in London who is preparing to scale their machine learning capability. They need an experienced MLOps Engineer who can take models from development into a reliable production environment and keep everything running smoothly throughout the lifecycle.

What you will be working on

• Building, improving and maintaining the end to end ML pipeline from data ingestion through to deployment.
• Working closely with Data Scientists to turn notebooks and prototypes into production ready workloads.
• Setting up CI and CD for model training, testing and deployment using Python based toolsets.
• Managing model performance, drift monitoring and regular retraining workflows.
• Creating infrastructure using Terraform or similar, ideally within Azure.
• Maintaining containerised workloads using Docker and Kubernetes.
• Supporting observability by putting in place clear logging, alerting and metrics for ML services.
• Ensuring best practice for version control, experiment tracking and model registry.
• Troubleshooting production issues and improving stability and performance across the ML platform.

Key skills

• Strong Python for production environments.
• Solid experience with Azure for ML workloads including compute, networking and storage.
• Good knowledge of MLOps tools such as MLflow, Azure ML, Kubeflow or similar.
• Strong background in CI and CD using GitHub Actions, Azure DevOps or similar.
• Experience working in containerised environments using Docker and Kubernetes.
• Clear understanding of how to support Data Scientists and engineering teams through the model lifecycle.

Useful to have

• Experience working on large scale data or ML projects.
• Good understanding of data engineering fundamentals.
• Experience working in regulated or high availability environments.

If you're an experienced MLOps Engineer feel free to apply or send your C.V to nokakou@revoco-talent.co.uk

MLOps Engineer