Course Wiki

FastAPI

The Python web framework ML Zoomcamp uses to expose trained models as prediction services and AI Dev Tools Zoomcamp uses for API backends.

In Machine Learning Zoomcamp, deployment starts when a trained model becomes a web service. FastAPI wraps the pickled model in a predict endpoint: the service loads the model at startup, receives JSON feature payloads, and returns predictions. The course chose it for its automatic validation (type hints via pydantic) and built-in interactive docs.

The service then gets packaged into a Docker image and shipped to the cloud, which is the module-to-module arc of the course. AI Dev Tools Zoomcamp reuses FastAPI as the backend framework for its full-stack app, implemented from an OpenAPI contract with tests.

Taught in

• Machine Learning Zoomcamp

  • Module 5: Deploying Machine Learning Models
    • Explore more

• AI Dev Tools Zoomcamp

  • Module 2: Build and Ship an AI-Assisted Full-Stack App
    • Build and Ship an AI-Assisted Full-Stack App ## Related concepts

• Docker

• Serverless Deployment

• Model Deployment

• OpenAPI Contract

• Kubernetes