AdvancedContent in Spanish
MLOps: deploying and operating models
A model that only lives in a notebook adds no value: the value appears when it serves predictions in production reliably and reproducibly. In this course you will learn to take machine learning models to production from start to finish: you will version data and models with DVC and MLflow, package your code in Docker containers, serve inference with FastAPI, set up CI/CD pipelines, and monitor drift to decide when to retrain. It is the fifth and last course of the Professional Certificate in Data Science, the one that turns your models into production systems.
9 lessons 12 hours 900 XP
Start the courseSyllabus
Module 1 · The ML lifecycle in production
- 1From notebook to production: what MLOps is45 min
Module 2 · Versioning data and experiments
- 2Versioning datasets with DVC55 min
- 3Experiment tracking with MLflow Tracking60 min
- 4Packaging and registering models with MLflow55 min
Module 3 · Serving and containerizing models
- 5Serving models with a REST API in FastAPI65 min
- 6Containerizing the service with Docker60 min
- 7Batch inference at scale50 min
Module 4 · Automation, monitoring and governance
- 8CI/CD for machine learning60 min
- 9Monitoring, drift and retraining55 min