AI & ML Academy - ML Engineering in Production (MLOps)

Welcome to the AI & ML Academy (AIA) - ML Engineering in Production (MLOps)!

Machine learning Operations (MLOps) applies DevOps principles and techniques to machine learning projects and help in efficiently scaling your project from experimentation to production. This module illustrates key DevOps concepts such as source control, automation, and CI/CD to build an end-to-end MLOps solution using popular tools such as Azure DevOps and GitHub Actions.

Getting Started - A Journey, not a Destination!

  • Introduction to MLOps - this is a great resource if you are new to MLOps and looking for a learning resource for a decent understanding of the process.
  • MLOps Best Practices - this is a one-pager infographic that walks through 5 best practices to optimize your MLOps lifecycle on Azure.
  • Comparisons of MSFT and Open Source Tools
    • Overview of Azure DevOps and GitHub Actions Comparing Azure DevOps and GitHub Actions
    • Overview of Kubeflow and MLflow Comparing Kubeflow and MLflow

MLOps with Azure DevOps

  • Azure MLOps (v2) Solution Accelerator - This accelerator is intended to serve as the starting point for MLOps implementation in Azure using Azure DevOps.
  • MLOps From Scratch hack using Azure DevOps - This challenge-based hack will introduce you to the Azure DevOps tooling for MLOps and helps you get hands-on experience with creating end-to-end Build and Release pipelines to continuously train and deploy your ML model to production.

MLOps with GitHub Actions

MLOps with Kubeflow

MLOps with mlFlow