A new book has been published about building machine learning systems.
Machine Learning Engineering by Andriy Burkov, is a complete guide to building, testing, deploying and maintaining machine learning models in real world settings. Very kindly the book is made available on a "read first, buy later" basis. I'm currently working my way though it but so far have been really impressed. This is a rapidly evolving field and one that has yet to establish many best practices or standardised approaches. I've done quite a lot of work in this areas and everything I've read so far i've either agreed with, or it has been useful information for me.
Rather than focusing on specific algorithms or packages, the book looks at ways of solving problems. It also includes practical code snippets in both R and Python. I think this is a really useful contribution to the field.