Machine Learning Engineering with Python: Manage the production life cycle of machine learning models using MLOps with practical examples

There are a lot of books out there that walk you through the steps of putting together a complex ML model using ideal data in a closed setting. This is not one of those books. ML engineering with Python is instead a comprehensive guide to the way machine learning works in practice at most companies. The book does a great job of explaining the MLops tools that almost all businesses today rely on to train, deploy, serve, and iterate on models. In my opinion, the concepts in this book are far more valuable than understanding how to use specific ML frameworks to solve problems. Simply understanding that these tools exist, and knowing how they are used will give engineers a leg up, and lead to more revenue generating impact than any gold medal kaggle model could produce on its own.

Machine Learning engineering with Python - I would highly recommended this book for intermediate level data scientist/ ML engineers who has learned the modelling skills and want to take it forward to successfully implement the solution with advanced software engineering techniques. Author rightly understands the current gap in understanding on implementation techniques in the market and addresses the same with multiple end to end example of real-time/batch/forecasting etc. Book focuses on many important areas like designing, tracking and versioning of code, model and data (data drift) using the tools needs at each stage - model training, model re-training when drift is detected, saving the feature transformation, automating hyper parameters with Optuna and HyperOpt and pipelines and packaging it properly for testing, logging and error handling. Chapter 5 : Deployment Pattern & Chapter 6 : Scaling up stood out for me where author described various implementation patterns and perform vertical/horizontal scaling. This was a new learning for me. Additionally there was great use of pictures, tables and architecture diagrams that was very helpful. Scope of Improvement : 1. Since Author focused deployment only on AWS, readers from Azure/Google Cloud may feel left out. 2. End to end examples didn't feel end to end from the perspective of code. New people coming into the field won't be able to follow end to end examples. I felt, I problem statement and detailed implementation would be a great addition in the next version.

This book will help you fill the gaps in your understanding of machine learning engineering and machine learning development process. Models in production constantly suffer from data drift, from the need to retrain and maintain the models in the pipelines. The authors provide a comprehensive overview of the modern approaches and give examples of real life solutions. You will find examples with Apache Spark and serverless architecture as well as AWS. What I liked the most was the dataset and code examples in the github repo that goes together with the book. The examples are given in the python notebook files, starting from simple solutions as detecting anomalies and to specific and more narrow examples of how to continuously retrain a model in the serverless cloud. This book will definitely be interesting for engineers who start deploying their models in production and want to make this process work the best way for their business.

Book is a fine introduction to machine learning using Python and covers all the stages of ML with lots of practical exercises. Last chapter covers end to end examples with usecase and it is very helpful. Recommend this book as it is a excellent guide to starting machine learning

The first chapter provided a really thorough breakdown surrounding the different data-oriented roles. It was clear to me the responsibilities of each of the roles. What was most helpful was navigating the table of contents, so that I was able to go to the necessary sections I was interested in studying. The examples of where we can apply ML and the System Design breakdown was most helpful for me to understand more of the systems engineering concepts. One thing I would recommend would be to have some Q&A for the reader to test their knowledge or understanding of the concepts and when/where to apply them. I would also be interested in hearing the author's perspective on how they think the world of ML will change, mainly the technologies being leveraged (I liked the AutoML example).

Very well written. It has a good introduction, meaty content with great examples. The production life cycle of MLE models was explained with great clarity!!! I love how you are given two end-to-end examples in the last two chapters. You can use them as templates for future projects.

Nicely explained. The examples and code snippets make the book a handy guide and reference. Nicely put together. All stages of ML and underlying concepts are covered in detail.

The concepts explained are very pragmatic and in a concise way. There are a lot of practical examples which makes it easier to understand the concepts.

I have been working my way through this book for the last couple of weeks. It's very well written, the examples are practical and realistic, and the advice is very useful. It's an up to date take on a rapidly evolving field. Highly recommended!

Lots of great insights, clearly explained and with very practical examples. Highly recommended.