Course Wiki

Flask

The lightweight web framework MLOps Zoomcamp uses to turn a trained model into a prediction web service.

Module 4’s web-service deployment is a Flask application: it loads the NY Taxi duration model from the model registry at startup, exposes a /predict endpoint that takes JSON ride features, and returns the prediction. The service is packaged with a Dockerfile and run with gunicorn inside Docker.

Flask’s minimalism is the point — the deployment lessons focus on the surrounding machinery (registry access, containerization, testing with docker-compose) rather than framework features. The same pattern is what Machine Learning Zoomcamp’s FastAPI module mirrors with type-safe request validation.

Taught in

• Machine Learning Zoomcamp

  • Module 5: Deploying Machine Learning Models
    • Intro / Session overview
    • Web services: introduction to Flask
    • Serving the churn model with Flask
    • Python virtual environment: Pipenv
    • Environment management: Docker
    • Deployment to the cloud: AWS Elastic Beanstalk (optional)
    • Explore more
  • Module 10: Kubernetes and TensorFlow Serving
    • Creating a pre-processing service
    • Deploying TensorFlow models to Kubernetes

• LLM Zoomcamp

  • Module 7: End-to-End Project
    • End-to-End Project Example
    • Interface and Ingestion Pipeline
    • Monitoring and Containerization
    • Summary and Closing Remarks

• MLOps Zoomcamp

  • Module 4: Model Deployment
    • Web-services: Deploying models with Flask and Docker ## Related concepts

• Model Deployment

• Model Registry

• Docker

• FastAPI

• Model Monitoring