Machine Learning (ML) · Advanced level

Advanced level · Module 4: Deploying Machine Learning Models · Lesson 3 of 4

Batch and Real-Time Deployment

Choose between scheduled batch predictions and on-demand APIs based on how predictions are used.

In this lesson you will

  • Compare batch and real-time prediction
  • Implement a scheduled batch job

Key concepts

batch scoringreal-time inferencelatencyscheduling

The lesson

Batch deployment runs predictions on a schedule — for example every month for all markets — and stores the results. It is simple, robust and works well when decisions are periodic.

Real-time deployment answers requests on demand, needed when predictions depend on fresh input, such as a loan officer checking an application. It requires a service to be running and monitored.

Many organisations need only batch predictions. Choose the simpler option unless the decision truly requires immediate answers.

Step by step

  1. List how the price forecasts will be used.
  2. Write a batch script that produces a monthly report.
  3. Schedule it with cron or Task Scheduler.
  4. Compare its operating effort with the API.

Practice activity

Implement the monthly batch forecast and recommend batch or real-time deployment for the early-warning service.

Expected result: A working batch job and a justified deployment recommendation.

Check your understanding

  1. When is batch deployment enough?

    Show answer

    When decisions are made periodically and do not need instant predictions.

  2. What does real-time deployment require?

    Show answer

    A running, monitored service that answers requests on demand.

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.