Master's year 1 · Full module

End-to-end MLOps

Take a model from notebook to production: experiment tracking, containers, orchestration and deployment, with a final ticket-classification project.

Learning objectives

• Structure a reproducible ML project
• Track experiments and version models with MLflow
• Containerize and deploy a model with Docker and Kubernetes
• Automate an end-to-end pipeline

Course outline Free preview

1. Model lifecycle and MLOps culture
2. Reproducibility: environments, data, code
3. Experiment tracking with MLflow
4. Docker, Kubernetes and prediction APIs
5. Orchestration and monitoring
6. Final project: ticket classifier