Machine Learning (ML) · Advanced level

Advanced level · Module 6: Machine Learning Capstone · Lesson 4 of 5

Deployment and Monitoring

Deploy the service, track it and demonstrate drift detection and a retraining response.

In this lesson you will

  • Deploy the forecasting service reproducibly
  • Demonstrate monitoring and retraining

Key concepts

API and batch deploymentexperiment trackingdrift simulationrunbook

The lesson

Deploy the forecasts as both a monthly batch report and an API, packaged to run from the README.

Track experiments and register the model. Simulate new months including a shock, detect it through monitoring and carry out the retraining procedure.

Write the runbook: schedules, alerts, retraining, rollback and the owner.

Step by step

  1. Package and deploy the service.
  2. Register the model and set up monitoring.
  3. Run the drift simulation and respond.
  4. Write the runbook.

Practice activity

Complete milestone 4: deployment with demonstrated monitoring.

Expected result: A running service whose monitoring detects the simulated shock, with a documented response.

Check your understanding

  1. What should the runbook name?

    Show answer

    Schedules, alerts, retraining and rollback steps and the model owner.

  2. Why simulate drift before go-live?

    Show answer

    To prove that monitoring and response actually work.

Taught live by an instructor — online or in person. Free course lessons are for self-study and do not lead to a certificate — see registration options.