Machine Learning Model Serving Patterns and Best Practices

Become a successful machine learning professional by effortlessly deploying machine learning models to production and implementing cloud-based machine learning models for widespread organizational use

Key Features
Learn best practices about bringing your models to production
Explore the tools available for serving ML models and the differences between them
Understand state-of-the-art monitoring approaches for model serving implementations
Book Description
Serving patterns enable data science and ML teams to bring their models to production. Most ML models are not deployed for consumers, so ML engineers need to know the critical steps for how to serve an ML model.

This book will cover the whole process, from the basic concepts like stateful and stateless serving to the advantages and challenges of each. Batch, real-time, and continuous model serving techniques will also be covered in detail. Later chapters will give detailed examples of keyed prediction techniques and ensemble patterns. Valuable associated technologies like TensorFlow severing, BentoML, and RayServe will also be discussed, making sure that you have a good understanding of the most important methods and techniques in model serving. Later, you'll cover topics such as monitoring and performance optimization, as well as strategies for managing model drift and handling updates and versioning. The book will provide practical guidance and best practices for ensuring that your model serving pipeline is robust, scalable, and reliable. Additionally, this book will explore the use of cloud-based platforms and services for model serving using AWS SageMaker with the help of detailed examples.

By the end of this book, you'll be able to save and serve your model using state-of-the-art techniques.

What you will learn
Explore specific patterns in model serving that are crucial for every data science professional
Understand how to serve machine learning models using different techniques
Discover the various approaches to stateless serving
Implement advanced techniques for batch and streaming model serving
Get to grips with the fundamental concepts in continued model evaluation
Serve machine learning models using a fully managed AWS Sagemaker cloud solution
Who this book is for
This book is for machine learning engineers and data scientists who want to bring their models into production. Those who are familiar with machine learning and have experience of using machine learning techniques but are looking for options and strategies to bring their models to production will find great value in this book. Working knowledge of Python programming is a must to get started.

Table of Contents
Introducing Model Serving
Introducing Model Serving Patterns
Stateless Model Serving
Continuous Model Evaluation
Keyed Prediction
Batch Model Serving Pattern
Online Learning Model Serving
Two-Phase Model Pattern
Pipeline Pattern Model Serving
Ensemble Model Serving Pattern
Business Logic Pattern
Exploring Tensorflow Serving
Using Ray Serve
Using BentoML
Serving ML Models using a Fully Managed Cloud Solution