This repository provides a self-paced course on Serverless Machine Learning, targeting individuals with Python and ML fundamentals who want to build end-to-end AI-enabled prediction services without deep infrastructure expertise. It emphasizes practical MLOps principles and operationalizing models using serverless architectures.

How It Works

The course leverages a serverless approach, abstracting away infrastructure management. It guides users through building batch and real-time prediction services using Python pipelines orchestrated via GitHub Actions. Key components include a serverless feature store and model registry (Hopsworks) for managing ML artifacts and enabling feature engineering, model training, and inference pipelines.

Quick Start & Requirements

• Install/Run: No specific installation command; course content is delivered via lectures, slides, and labs.
• Prerequisites: Python (with Pandas), a GitHub account, and a Hopsworks.ai account. A Python IDE (Jupyter/Colaboratory) is recommended.
• Resources: GitHub Actions provides a free tier (500 MB storage, 2,000 minutes compute). Hopsworks.ai offers a free tier (10 GB storage).
• Links:
  • Course Overview: https://github.com/featurestoreorg/serverless-ml-course
  • GitHub Actions Billing: https://docs.github.com/en/billing/managing-billing-for-github-actions/about-billing-for-github-actions

Highlighted Details

• Focuses on building production-ready ML systems beyond notebooks.
• Covers MLOps essentials: versioning, testing, data validation, and operations.
• Demonstrates building serverless UIs for prediction services using Gradio, GitHub Pages, or Streamlit.
• Utilizes free tiers of GitHub Actions and Hopsworks.ai for cost-effective development.

Maintenance & Community

The course is associated with the Feature Store Org community. Links to related resources like Awesome MLOps, MLOps Zoomcamp, and Full Stack Deep Learning are provided.

Licensing & Compatibility

The repository itself does not specify a license. The course content is free to use, leveraging services with their own terms and free tiers.

Limitations & Caveats

The course assumes familiarity with ML concepts and Python programming. While aiming to abstract infrastructure, a basic understanding of GitHub and Python scripting is necessary for practical application.