Learning path
MLOps: Deploying Models
Model APIs with FastAPI and Docker, cloud ML platforms, and monitoring for drift, with animated films and runnable Python. Part of the Data Scientist premium track.
-
1
Serving a Model with FastAPI and Docker
Take a trained scikit-learn model from a notebook to a real API: save it with joblib, serve predictions with FastAPI and input validation, and package it in a…
-
2
Cloud ML: AWS SageMaker and Azure Machine Learning
How teams train and deploy models in the cloud: AWS SageMaker and Azure ML compared, compute choices (CPU vs GPU), experiment tracking with MLflow, model registries and managed…
-
3
Monitoring Models in Production
Models decay. Detect data drift with PSI and statistical tests, track prediction and business metrics, set alerts, and decide when to retrain, with runnable Python.