Practical MLOps: Operationalizing Machine Learning Models

by Noah Gift, Alfredo Deza

Getting your models into production is the fundamental challenge of machine learning. MLOps offers a set of proven principles aimed at solving this problem in a reliable and automated way. This insightful guide takes you through what MLOps is (and how it differs from DevOps) and shows you how to put it into practice to operationalize your machine learning models.

Current and aspiring machine learning engineers--or anyone familiar with data science and Python--will build a foundation in MLOps tools and methods (along with AutoML and monitoring and logging), then learn how to implement them in AWS, Microsoft Azure, and Google Cloud. The faster you deliver a machine learning system that works, the faster you can focus on the business problems you're trying to crack. This book gives you a head start.

You'll discover how to:


Apply DevOps best practices to machine learning
Build production machine learning systems and maintain them
Monitor, instrument, load-test, and operationalize machine learning systems
Choose the correct MLOps tools for a given machine learning task
Run machine learning models on a variety of platforms and devices, including mobile phones and specialized hardware

Genres: Computer Science, Programming, Artificial Intelligence, Software, Technical, Technology

462 pages, Kindle Edition

Published September 14, 2021

About the author

Noah Gift is lecturer and consultant at UC Davis Graduate School of Management in the MSBA program, Northwestern's Master of Data Science program, and UC Berkeley Data Science program. He is also the founder of Pragmatic AI Labs. At Pragmatic AI Labs he provides consulting and training on Machine Learning and AI, and also develops AI SaaS products. At UC Davis, he is teaching graduate machine learning and consulting on Machine Learning and Cloud Architecture for students and faculty. He has published close to 100 technical publications including two books on subjects ranging from Pragmatic AI: An Introduction to Cloud-Based Machine Learning, First Edition to DevOps. He is also a certified AWS Solutions Architect and has an MBA from UC Davis, a M.S. in Computer Information Systems from Cal State Los Angeles, and a B.S. in Nutritional Science from Cal Poly San Luis Obispo. Professionally, Noah has approximately 20 years’ experience programming in Python and is a member of the Python Software Foundation. He has worked in roles ranging from CTO, General Manager, Consulting CTO and Cloud Architect. This experience has been with a wide variety of companies including ABC, Caltech, Sony Imageworks, Disney Feature Animation, Weta Digital, AT&T, Turner Studios and Linden Lab. In the last ten years, he has been responsible for shipping many new products at multiple companies that generated millions of dollars of revenue and had global scale. Currently he is consulting startups and other companies, on Machine Learning, Cloud Architecture and CTO level consulting via Pragmatic AI Labs

Reviews

[Vaidas · Rating 2 out of 5]

Two stars because this book's information is useful and encourages the right mindset.
However, the title should be changed to "Starting Mlops: how to start putting your models to production".

I picked up this book to learn patterns and best practices of how to do a world-class mlops. This book did not deliver on that - there are good ideas, basically a reiteration of DevOps ideas in ML context, but there are no practical examples of really powerful Mlops. The examples for cloud functions demonstrate *copying* your code to GUI, which for me is anti-mlops.

If you are new to ML and DevOps - you might find this book useful. However, if you worked on these problems before, I don't think you will learn much from it.

[Aileen · Rating 3 out of 5]

A great reference for beginners but minimal new usable concepts for those a little more experienced.

[Federico · Rating 4 out of 5]

The authors provide some concrete explanation of what can be meant with MLOps, without any mysticism. Finally I feel I understood!

Most of the pages are spent on rather concrete examples for specific hosting providers (all of the usual suspects among the hyperscalers). So the specific recipes are not cloud-agnostic, but the book overall is sufficiently balanced.

[Amirali · Rating 1 out of 5]

This book's target audience is beginners with no idea what Mlops is and what to see how a running non productionalized application can be. Also, the amount of non-relevant examples is very annoying since they make it hard to skim and only grasp new concepts you are unfamiliar with.

[mahesh · Rating 3 out of 5]

I was expecting some best practices that can be used in MLOPS, However, this book is purely for beginners who don't know what is MLOPS. The information in it was not new to me.
Read this book if you are new to MLOPS.

[Aaron · Rating 3 out of 5]

Software guide for cloud providers but gets bogged down by anecdotes and insists upon pre made models could have been more technical in its approach. More emphasis on different software was maybe needed and building things from scratch. Too much emphasis on really nonsense things like home office decor and some long winded anecdotes in the final chapters

[Matthew Mechtly · Rating 2 out of 5]

This is a textbook for people with little to no experience in MLOps or DevOps. This does not match the rigor, organization, or thoughtfulness that usually comes from an O'Reilly book. In fact, this book was so elementary and disorganized at time that I stopped reading entire sections and ultimately couldn't finish the book

However, if you are approaching MLOps as an absolute beginner, this could be a decent resource, though there are almost certainly better options.