Micro Free Course — 100% Free Learning
All lessons in this module are free to learn. Sign in with Google to save your progress.
Live MLOps Cases
Six hands-on end-to-end walkthroughs: reproducible training with MLflow, containerized FastAPI deployment, drift detection setup, debugging a model that broke in prod, designing a retraining pipeline, and running a model A/B test.
Module Content
Case 1 — Build a Reproducible Training Pipeline with MLflow
Take notebook-trained code and turn it into a versioned, reproducible MLflow pipeline.
Case 2 — Deploy a Model as a Containerized FastAPI Service
From trained model to running endpoint in production. Full deploy.
Case 3 — Set Up Drift Detection on Production Predictions
Predictions go silently bad without monitoring. Build the drift detection that catches it.
Case 4 — Diagnose a Model That Worked Yesterday but Not Today
Predictions broke overnight. Walk through the diagnostic process to find the cause fast.
Case 5 — Design a Retraining Pipeline for a Fraud Model
Daily retraining is overkill; quarterly is too rare. Design the right cadence and pipeline.
Case 6 — Run an A/B Test Between Two Model Versions
A new model passes validation. Don't deploy it blindly — A/B test it against the current production model.