[Book Review] Amazon Bedrock AgentCore Practical Introduction ─ A Book for Building and Operating AI Agents on AWS

Introduction

Hello, I'm Jinno from the consulting department, and I'm a huge fan of Amazon Bedrock AgentCore.

This time, I'd like to introduce the book Amazon Bedrock AgentCore: A Practical Introduction, which will be released on May 29, 2026!

https://www.sbcr.jp/product/4815641238/

I was fortunate enough to be given the opportunity to read the manuscript as a technical reviewer while the book was still being written. Throughout the review process, I kept thinking to myself, "This is an incredible book — it's packed with knowledge about AgentCore, which I love so much."

This is my first time writing a book review blog post, but as someone who has read through every chapter, I hope to share my honest impressions!

Book Overview

The three authors are also the authors of the bestselling previous work, Amazon Bedrock: An Introduction to Generative AI App Development.

They are individuals who are at the forefront of the generative AI and AI agent space, and I have always looked up to them. When I first read Amazon Bedrock 生成AIアプリ開発入門, I knew nothing about Bedrock and was able to catch up in an easy-to-understand way through the book, so I had come to admire all three of them on my own. I never dreamed I would get to review the follow-up book by these wonderful people, and I'm truly grateful that such a day has come.

Target Readers

The cover of the book lists the following as "target readers":

• Those who want to learn how to build efficiently on AWS
• Those who want to take on the challenge of developing high-performance AI agents
• Those who want to learn through hands-on experience
• Those who want to learn systematically from understanding the overall structure

The primary audience is IT engineers and developers, but the book is also structured with consideration for those who have become interested in in-house development through AI. There are many sample code examples, and since tools like ChatGPT have web search features, you can leverage AI as you go, allowing even those who don't normally develop software to steadily work through the material with understanding. Actually building things is the best way to learn.

A sample code repository is also provided, so make great use of it as a reference when building things yourself!

https://github.com/minorun365/agentcore-book

Table of Contents

The book is structured into 4 parts, 16 chapters, plus appendices.

Part 1: Fundamentals

• Chapter 1: Generative AI Basics and Introduction to Amazon Bedrock
• Chapter 2: Introduction to AI Agents

Part 2: Strands Agents

• Chapter 3: Introduction to Strands Agents
• Chapter 4: [Hands-On] Let's Build a Research Agent

Part 3: AgentCore

• Chapter 5: Overview of AgentCore and Its Main Feature, "Runtime"
• Chapter 6: "Memory" for Managing Memories
• Chapter 7: "Identity" for Controlling External Authentication
• Chapter 8: "Gateway" for Bundling Tools
• Chapter 9: "Policy" for Controlling Tool Usage
• Chapter 10: "Built-in Tools" Unique to the Cloud
• Chapter 11: "Observability" for Visualizing Operational Status
• Chapter 12: "Evaluation" for Automated Quality Assessment
• Chapter 13: [Hands-On] Let's Build a Full-Stack Agent

Part 4: Advanced Applications

• Chapter 14: Leveraging Internal Data for Agents with RAG
• Chapter 15: [Hands-On] Let's Build an Ambient Agent with CDK
• Chapter 16: Introducing AI Agents into Business Operations

Appendices

• Appendix 1: Setting Up an AWS Account
• Appendix 2: Setting Up a Development Environment
• Appendix 3: Bedrock Service Quotas

Part 3, the AgentCore section, accounts for approximately 40% of the entire book. True to the title "AgentCore Practical Introduction," a notable feature is that each of AgentCore's functions is given its own dedicated chapter.

My Personal Highlights!

Full Color with Abundant Diagrams — Incredibly Easy to Understand

The first thing you notice when reading the book is the sheer number and quality of the diagrams.

AgentCore has a wide range of features, and organizing the individual functions themselves as well as their relationships can be quite challenging...

The book carefully explains these concepts using original diagrams, allowing you to intuitively grasp how each feature works and how they interact with one another.

Even parts that are difficult to understand from official documentation alone become clear — you find yourself thinking "Ah, so that's what it means!" — because diagrams and explanations are paired together, which is incredibly helpful.

Truly Tracking the Very Latest Updates

There are many points where I found myself thinking, "They've even incorporated that update..." It was simply impressive.

AgentCore and Strands Agents are services that are evolving rapidly, but this book continues to incorporate the latest updates. New features such as AgentCore Payments (payment functionality from agents), AgentCore Optimization (continuous improvement based on evaluation results), HTTP target support for Gateway, and the OBO (On Behalf Of) flow for Identity are reflected in columns and the main text, and I could feel the dedication of the authors.

Change in this field is truly fast-paced, so the fact that things are systematically compiled as of May 2026 is itself highly valuable, and no other book is keeping up with developments to this degree.

Hands-On That Goes All the Way to SaaS Integration

This is my personal top recommendation!

In Chapter 13, you get to experience a truly "full-stack" hands-on where you build a calendar assistant that integrates with the Google Calendar API via OAuth, manage credentials with AgentCore Identity, and deploy a frontend using Next.js + Amplify.

When actually trying to introduce AI agents in a business setting, frontend implementation and authentication integration with external SaaS products are unavoidable challenges. Hands-on content that tackles these head-on is something you rarely see in other books or web resources. Three-legged OAuth (3LO) using Identity is admittedly complex, but this hands-on breaks it down in a very accessible way, making it possible to understand and build it yourself.

The frontend is also built with a modern UI, making it something you could actually incorporate into real business work — which is great.

Highlights by Section

Part 1: Fundamentals — Grasping the Big Picture of AI Agents

Chapter 1 starts from the basics of generative AI and Amazon Bedrock, and even includes a hands-on session using the Converse API and ConverseStream API. The structure is welcoming even to those encountering Bedrock for the first time.

Chapter 2's Introduction to AI Agents covers a wide range of topics including the ReAct pattern, agent design patterns, and commonly appearing standard protocols like MCP and A2A. With many diagrams and clear explanations, just this chapter alone gives you a pretty clear overall picture of "what is an AI agent."

Part 2: Strands Agents — Understanding the Framework

You can learn step by step, from the basic usage of Strands Agents all the way to implementing multi-agent patterns.

The Chapter 4 hands-on has you build a research agent using the Swarm pattern (a pattern where multiple agents cooperate). The mechanism by which five agents — PlanAgent, WhatsNewSearchAgent, RetrievalAgent, ReviewAgent, and OrchestratorAgent — work cooperatively, communicating with each other to produce output, feels futuristic and fascinating. Practical examples of Swarm like this are rare, making it very valuable as a reference. Using Opus, Sonnet, and Haiku models according to each agent's role, and configuring prompt caching to keep costs down — these feel like genuinely practical points.

Part 3: AgentCore — The Core of the Book

This is the thickest part of the book, explaining each of AgentCore's 8 functions in a dedicated chapter.

Each chapter follows a consistent structure: concept explanation → code examples → pricing. The fact that pricing is also addressed in each chapter is a welcome feature, as it prevents the worry of "I'd like to try it, but what about the cost..."

As the book itself discusses, AgentCore has many features, and it's not necessary to use all of them — it's best to adopt only the features you need, à la carte style. That's why it's important to understand each feature individually, and having them explained carefully one by one is really helpful. Even features like Gateway, whose benefits might not be immediately obvious, are explained in detail, so make sure not to miss them.

Part 4: Advanced Applications — A Path to Business Adoption

Chapter 14 on RAG dives into leveraging internal data using Bedrock Knowledge Base, while Chapter 15 covers building an ambient agent using CDK, taking the content into increasingly practical territory.

For document search, AI agents + RAG continues to be in demand, and the book carefully explains the configurations you'll want to have under your belt. It also touches on query decomposition and reranking methods to improve search accuracy with Knowledge Bases, with introductions that allow you to implement them as needed — which is greatly appreciated. After all, building something and stopping there isn't enough; accuracy verification and improvement are also essential.

The "3-Step Recommended RAG Implementation Process" introduced in a column is also practical, laying out a clear progression: first build simply with S3 Vectors → introduce Advanced RAG incrementally → measure accuracy quantitatively through evaluation. Since it's easy to feel lost about where to start with RAG, what matters first is trying it out quickly and inexpensively to test its effectiveness.

The Chapter 15 hands-on for the ambient agent is also a highlight. Simply by uploading a receipt image to S3, the agent automatically handles everything from image analysis and expense categorization to sending approval request emails, and even recording in Confluence after approval. Rather than instructing via chat, this is something that runs autonomously in the background triggered by events, and since SaaS integration is something you often use in real work, it's very useful as a reference.

This hands-on also uses AWS CDK to define infrastructure as code, deploying resources such as AgentCore Runtime, S3, DynamoDB, and Lambda together. IaC is indispensable for production deployment, and being able to reuse infrastructure definitions you've built once in other projects is a major benefit. Since the book explains CDK fundamentals (such as the App / Stack / Construct hierarchy) from the ground up, even CDK beginners can tackle it with confidence.

Chapter 16 is written from an organizational and business perspective on "how to introduce AI agents into business operations." Personally, this chapter resonated with me more than any other.

In my own work supporting customers with AI agent adoption, it's not uncommon to start from a situation where "what AI agents can even do" is still vague. The book introduces a way of thinking about AI agents not as automation tools like RPA, but as "a talented mid-career hire." If you organize the rules and procedures of the work as prompts and documents, and set up the necessary access permissions to the right tools, the agent will appropriately get the job done. This analogy is easy to understand and something I want to keep in mind at all times.

The book also emphasizes the importance of "narrowing your scope and showing something that works first." Applications leveraging LLMs involve uncertainty, so trying to build a large-scale system all at once leads to struggle with the gap between expected and actual outputs. Build something that works at 50% completion and show it to stakeholders; accumulate "small wins" while getting a feel for how to leverage AI, and gradually make your AI utilization goals more concrete. This approach strongly resonates with my own experience.

And another thing that struck me was that many cases of failed adoption are not due to technical challenges, but rather the problem that "nobody uses it." Embedding it in users' daily workflows, making it possible to invoke the agent from within Slack, email, and other familiar UI surfaces — it would truly be a shame if something you put effort into building went unused...!

I sincerely hope both engineers and non-engineers alike will read this chapter.

Who This Book Is Recommended For

• Those who want to build AI agents on AWS but don't know where to start
• Those who want to learn Strands Agents and AgentCore systematically
• Those who want to acquire knowledge that covers not just building AI agents but also operating them
• Those who want to learn through hands-on practice
• Those who are in a position to propose AI agent adoption to clients

The content spans broadly from theoretical learning to actual code implementation, so while the primary audience is engineers, even non-engineers can dive in by leveraging generative AI. If you've decided "I'm going to build AI agents on AWS!" — this is the book for you!

Closing Thoughts

There is no other Japanese-language book that has compiled the large theme of building and operating AI agents centered around the feature-rich AgentCore in such a practical and systematic way. The structure — carefully explaining the basics of generative AI and each AgentCore function in dedicated chapters, while combining them together through hands-on sessions — creates a feeling of deepening understanding step by step, which made for a pleasant reading experience. It also served as a review for me personally, and it's a book I'll want to return to at any time. Lately, I've been reading it every morning on my commute by train.

2026 has the potential to be the inaugural year of building AI agents. For those thinking about introducing a self-built AI agent into their business operations during this turbulent year, this is the first book you should reach for.
Let's use this book to get up to speed and dive headfirst into building AI agents with AgentCore!!!
I'll also keep sharing the latest AgentCore information through my blog, not to be outdone!

https://dev.classmethod.jp/author/yjinn/

Thank you so much for reading this article all the way to the end! Nothing would make me happier than if this has conveyed even a little of the book's appeal!