Building Intelligent Cloud Applications: Develop Scalable Models Using Serverless Architectures with Azure

by Vicente Garc a, John Biggs, Jose Luis Calvo Salanova

Overview

Serverless computing is radically changing the way we build and deploy applications. With cloud providers running servers and managing machine resources, companies now can focus solely on the application’s business logic and functionality. This hands-on book shows experienced programmers how to build and deploy scalable machine learning and deep learning models using serverless architectures with Microsoft Azure.

You’ll learn step-by-step how to code machine learning into your projects using Python and pretrained models that include tools such as image recognition, speech recognition, and classification. You’ll also examine issues around deployment and continuous delivery, including scaling, security, and monitoring.

This book is divided into three parts with application examples woven throughout:

• Cloud-based development: Learn the basics of serverless computing with machine learning, Functions-as-a-Service (FaaS), and the use of APIs
• Adding intelligence: Create serverless applications using Azure Functions; learn how to use prebuilt machine learning and deep learning models
• Deployment and continuous delivery: Get up to speed with Azure Kubernetes Service, Azure Security Center, and Azure Monitoring

Product Details

ISBN-13:	9781492052326
Publisher:	O'Reilly Media, Incorporated
Publication date:	10/01/2019
Pages:	152
Product dimensions:	6.80(w) x 9.10(h) x 0.40(d)

About the Author

A seasoned Software Engineer, with solid experience in all facets of software and web development projects. Possess a Master’s Degree in Computer Science by University of Seville. Highly skilled in collaborating with all members of the organization and clients to achieve technical, business and financial objectives. Adept in streamlining processes, enhancing productivity levels, and introducing new emerging methodologies that enrich product development. Certified Scrum Master "Expert Level", he is a certification teacher and examiner at Scrum Manager, and also teaches software and web development, and project management subjects at Sevilla Universityand other private institutions.

John Biggs is an entrepreneur, consultant, writer, and maker. He spent fifteen years as an editor for Gizmodo, CrunchGear, and TechCrunch and has a deep background in hardware startups, 3D printing, and blockchain. His work has appeared in Men’s Health, Wired, and the New York Times. He runs the Technotopia podcast about a better future.

He has written five books including the best book on blogging, Bloggers Boot Camp, and a book about the most expensive timepiece ever made, Marie Antoinette’s Watch. He lives in Brooklyn, New York.

Table of Contents

Preface vii

Part I Cloud-Based Development

1 Machine Learning and Deep Learning Models in the Cloud 3

An Introduction to Machine Learning 3

An Introduction to Deep Learning 7

Neural Networks 7

Difficulties Defining Structure and Training Machine Learning Models 8

An Introduction to Serverless Machine Learning 10

Using Containers with Machine Learning Models 12

The Benefits of Serverless Computing for Machine Learning 13

2 Functions-as-a-Service and Event-Driven Programming 15

Software-as-a-Service, Cloud Computing, and Serverless 15

Microservices Architecture 18

The Rise of Functional Programming 19

Using Functions Instead of Objects 22

Asynchronous Programming 24

Serverless 25

Implementing Functions 28

Event-Driven Architecture 29

Implications of Real-Time Processing 30

Summary and Look Ahead 32

3 Serverless Application Programming Interfaces in Microsoft Azure 33

APIs in Serverless Platforms 33

An Introduction to Azure 34

Azure General Services 35

Part II Adding Intelligence

4 Getting Started with Microsoft Azure Functions 43

Azure Functions 43

Creating a Function App by Using Azure Portal 44

Local Development Environment 48

Creating a Functions Project Using Core Tools 52

Using Visual Studio Code 60

Debugging Python by Using Visual Studio Code 63

Azure Blob Storage 66

Creating a Storage Account 66

Transferring Blobs 68

Summary and Look Ahead 70

5 Using Machine Learning and Deep Learning Models 71

Azure Cognitive Services 71

Setting Up and Using a Service in Cognitive Services 72

Using a Cognitive Service from a Serverless Azure Functions Project 74

General Machine Learning Tools, Libraries, and Frameworks 75

Microsoft Cognitive Toolkit 75

ML.NET 77

Jupyter Notebook 78

TensorFlow 78

Keras 79

Scikit-learn 79

Cloud Machine Learning Solutions with Azure 81

Microsoft Machine Learning Studio 81

Azure Machine Learning Service 86

Part III Deployment and Continuous Delivery

6 Deployment and Scaling 93

Deployment Options 93

Azure DevOps 95

Docker 97

Azure Container Registry 98

Kubernetes 99

Machine Learning Tools for Kubernetes 102

Single-Container Machine Learning 105

Distributed Machine Learning with TFJob 108

7 Security 111

Azure Functions Authorization Levels 111

API Management 113

Azure Security 117

Operations 117

Applications 118

Storage 119

Network Layer Controls 119

Networking 120

8 Monitoring 123

Azure Monitor 123

Metrics and Logs 125

Data Sources 126

Application Insights 128

Insights for Containers (and VMs) 128

Log Analytics 129

Data Explorer Query Language 129

Alerts 130

Smart Groups 130

Autoscale and Metrics Alerts 131

Index 133