Course Wiki

MLOps Maturity Model

The ladder from manual, script-driven ML to fully automated training and deployment — the map MLOps Zoomcamp opens with.

Module 1 introduces the MLOps maturity model: level 0 is manual, notebook-driven ML where everything from training to deployment is scriptable at best; level 1 adds pipeline automation with experiment tracking and a training pipeline that can be re-run; level 2 adds CI/CD so training and deployment happen automatically on new data or new code.

The course positions every later module on this ladder — experiment tracking moves you off level 0, orchestration makes the pipeline repeatable, monitoring closes the loop. The NY Taxi duration-prediction dataset is the running example carried through all modules.

Taught in

• MLOps Zoomcamp

  • Module 1: Introduction
    • MLOps maturity model ## Related concepts

• Experiment Tracking

• Workflow Orchestration

• Model Deployment

• CI/CD

• Model Monitoring