Machine Learning on Kubernetes

Enhance your organization's collaboration and ML deployments by building a fast, reliable platform using available open source software that runs on Kubernetes

Key Features

• Build an open source machine learning platform on Kubernetes
• Develop a self-service machine learning workspace enabling MLOps
• Automate data pipelines and model deployment

Book Description

Ensuring that your machine learning models are suitable for every part of your organization is crucial to getting the maximum value for your team. By implementing MLOps with Kubernetes, data scientists, IT operations professionals, and data engineers will be able to collaborate and build ML solutions that create business value for your organization. This book enables teams to take a practical approach to working together to bring the software engineering discipline to the ML lifecycle.

You'll begin by understanding why MLOps is important and discover the different components of an ML project. Next, you'll design and build a practical end-to-end MLOps project that'll use the most popular OSS components. As you progress, you'll get to grips with the basics of MLOps and the value it can bring to your ML projects, as well as gain experience in building, configuring, and using an open source containerized ML platform on Kubernetes. Finally, you'll be able to prepare data, build and deploy models quickly, and automate tasks for an efficient ML pipeline using a common platform. The exercises in this book will help you get hands-on with using Kubernetes and integrating it with OSS such as JupyterHub, MLflow, and Airflow.

By the end of this book, you'll have learnt how to effectively build, train, and deploy an ML model using the ML platform you built.

What you will learn

• Understand what MLOps is and how to apply its different components
• Use open source software to build an MLOps platform on Kubernetes
• Build a complete project using the MLOps platform
• Improve on your organization's collaborative journey toward ML
• Discover how to use the platform as a data engineer, ML engineer, or data scientist
• Find out how to apply machine learning to solve your individual use cases

Who This Book Is For

This book is for data scientists, data engineers, IT platform owners, AI product owners, and data architects who want to build their own platform for ML development. Although this book starts with the basics, a good understanding of Python and Kubernetes, along with knowledge of the basic concepts of data science and data engineering will help you grasp the topics covered in a better way.

Table of Contents

1. Challenges in Machine Learning
2. Understanding MLOPs
3. Exploring Kubernetes
4. The Anatomy of a Machine Learning Platform
5. Data Engineering
6. Machine Learning Engineering
7. Model Deployment and Automation
8. Building a Complete ML Project Using the Platform
9. Building Your Data Pipeline
10. Building, Training, and Versioning Your Models
11. Deploying Models at Scale and Monitoring Them