MLOps Zoomcamp: a learning track from DevOps to MLOps
Added 2026-10-29.
Problem
Docker, CI and monitoring are familiar, but the ML part is missing: how to track experiments, version models and watch their quality in production.
Solution
"MLOps Zoomcamp: A Free 9-Week Course on Productionizing ML Services" from DataTalksClub — "master the fundamentals of MLOps, from training and experimentation to deployment and monitoring". It covers experiment tracking and a model registry (MLflow), orchestration, deployment, model monitoring, and a separate best practices module: pytest, docker-compose tests, localstack, IaC and CI/CD. There is no live cohort in 2026 — the course is self-paced.
Example
The modules that give a DevOps engineer the most:
| Module | What to take away |
|---|---|
| 02 — Experiment tracking | MLflow: experiments, model registry |
| 04 — Deployment | A model as a web service and batch processing |
| 05 — Monitoring | Data drift and model quality |
| 06 — Best practices | Tests, IaC and CI/CD for an ML project |
The outcome is one learning project in GitLab with CI, model deployment and monitoring.
Limitations
The repository doesn't state a license: use it for learning, don't copy the materials into the knowledge base. The cloud parts of the course can be replaced with local equivalents.
Source
How to apply it yourself
- Go through modules 02, 04, 05 and 06 on a single project.
- Publish the project to GitLab with CI and model monitoring.
- Mark it "Studied" once the project works end to end.