A model that only lives in a notebook isn't finished. MLOps is the practice of taking a trained model and turning it into something reliable that people and applications can actually use, then keeping it healthy after launch.
You'll start with the big picture of the MLOps lifecycle, then learn the building blocks: packaging your work with Docker and serving predictions through an API with FastAPI. The module closes with a full end-to-end project that brings in MLflow experiment tracking, DVC pipelines, and CI/CD deployment on the cloud.
This is the final module of the course. Once you've finished, head to the Projects section to put everything you've learned into practice.
An introduction to MLOps concepts and workflows, building a customer satisfaction prediction project with ZenML and a Streamlit app.
Containers versus virtual machines, core Docker commands, Docker Compose, volumes, and building and sharing your own images.
Build a Python API from scratch with path and query parameters, request bodies, and GET, POST, PUT, and DELETE endpoints, the foundation for serving a model.
An end-to-end project covering MLflow tracking, DVC pipelines, model registration, a Flask API in Docker, and CI/CD deployment on AWS.