A Iterative Approach for Developing Production-Ready Applications

Machine learning systems are both complex and unique. They are complex because they consist of many different components and involve many different parts in their development. They are unique because they depend on data, and the data varies significantly from use case to use case.

In this book, you will learn a holistic approach to designing machine learning systems that are reliable, scalable, maintainable, and adaptable to changing environments and business requirements. The author, Chip Huyen, co-founder of Claypot AI, examines every design decision—such as how to process and generate training data, which features to use, how often models should be retrained, and which metrics should be monitored—with an eye towards how it can help your system as a whole achieve its goals.

The iterative methodological framework presented in this book is based on real case studies, which are accompanied by a wealth of references. The book will help you successfully address needs such as:

• Data construction and selecting the appropriate metrics to solve a business problem
• Automating the processes of continuous development, evaluation, production deployment, and updating of models
• Developing a monitoring system to quickly detect and address problems that your models might face in production
• Designing a machine learning platform that can serve multiple different use cases
• Developing responsible machine learning systems

"This is simply the best book you can read on how to build, deploy, and scale machine learning models in a company to have the maximum impact." - Josh Wills, Software Engineer at WeaveGrid and former Director of Data Engineering at Slack

"In a thriving but chaotic ecosystem, this specific, comprehensive end-to-end view of machine learning serves as both a map and a compass: it’s a book that professionals inside and outside large tech companies should read." - Jacopo Tagliabue, Director of Artificial Intelligence at Coveo

Chip Huyen is the co-founder of Claypot AI, a real-time machine learning platform. Through her work at NVIDIA, Netflix, and Snorkel AI, she has helped some of the largest companies worldwide develop and deploy machine learning systems. Chip based this book on her lectures for CS 329S: Designing Machine Learning Systems, a course she teaches at Stanford University.