Courses Details and Modules

Why Serverless ML?


You should not need to be an expert in Kubernetes or cloud computing to build an end-to-end service that makes intelligent decisions with the help of a ML model. Serverless ML makes it easy to build a system that uses ML models to make predictions. You do not need to install, upgrade, or operate any systems. You only need to be able to write Python programs that can be scheduled to run as pipelines. The features and models your pipelines produce are managed by a serverless feature store / model registry. We will also show you how to build a UI for your prediction service by writing Python and some HTML.


Youtube Playlist
Zoom link

Module 00 (Optional)

Introductions, requirements & Machine Learning 101

Starting Date: Self Paced.
  • Lecture 0 - Why Serverless ML? link
  • Lecture 0 - Introduction to the course link
    • What we cover - what we don’t cover
  • Lecture 0 - Development Environment & Platforms link
  • Lecture 0 - Introduction to Machine Learning (ML 101) link

Module 01

How to write Pipelines in Python and run your first Prediction Service

Starting Date: Tuesday 27th September 2022 - 5PM CET
Lab Form
  • Feature Engineering in Pandas
  • What is a ML pipeline?
  • What is a ML pipeline?
  • Iris as a Prediction Service
  • Refactor notebooks into Python modules/functions and Notebooks
  • Github Actions
  • User Interface

Module 02

Data modeling and the Feature Store. The Credit-card fraud prediction service.

Release Date: 4th Oct 2022
Lab Form
  • Data modeling and the Data Warehouse / Feature Store
  • Feature Pipelines with synthetic data 
  • Feature Store

Module 03

Training Pipelines, Inference Pipelines, and the Model Registry.

Release Date: 11th Oct 2022
  • Transformations
  • Feature store Transformations vs Scikit-Learn Transformations
  • Model Registry
  • Get Batch Data
  • hsfs.get_feature_vector(...)

Module 04

Serverless User Interfaces for Machine Learning Systems.

Release Date: 24th Oct 2022
  • Effective Stakeholder Communication
  • Dashboards
  • Interactive User Interfaces
  • Application Integration

Module 05

MLOps Principles: Automated Testing, Versioning, Upgrades/Rollback

Release Date: 15th Nov 2022
  • Developing new versions of a feature and a model
  • Feature Logic tests: Pytest for Python functions outside notebooks
  • End-to-End tests with sample data
  • Data Quality tests with Great Expectations

Module 06

Real-time Machine Learning Systems

Release Date: 22 Dec 2022
  • Online Inference Pipelines
  • Training/Serving Skew
  • Model Deployments
  • KServe

Key Technologies 

Development environment

You can write, test, debug, and train your models in some Python IDE. We will focus on notebooks and Python programs. You can use Jupyter notebooks or Colabatory.


Github to manage your code, GitHub Actions to run your workflows, and Github Pages for your user interface for non-interactive applications. Github Actions offers a free tier of 500 MB and 2,000 minutes to run your pipelines.

Hopsworks has a free tier of 10 GB of storage.