Review: AI Agents in Action
If you’re trying to make sense of how to actually build AI agents, not just talk about them, AI Agents in Action might be for you.
About the author
Michael Lanham, Lead AI Developer at Brilliant Harvest, is a seasoned software and technology innovator with more than 25 years of industry experience. He is the author of 10 books, including Evolutionary Deep Learning.
Inside the book
Micheal Lanham’s latest book is not an abstract discussion about AI futures or lofty concepts. It’s a toolkit in print, aimed at developers and technical leaders who want to build real systems using language models, agent frameworks, and orchestration tools.
The book assumes you’ve already crossed the threshold into LLMs and want to move toward building structured, repeatable agentic systems. That means if you’re a CISO looking to understand the technical underpinnings of how LLM-driven agents could be deployed in enterprise workflows or how they might break, this is a solid resource to start that investigation. For developers and researchers, it’s a deep dive into the architectures, libraries, and techniques now forming the foundation of agent development.
One of the strengths of the book is how it layers complexity gradually. It starts with OpenAI’s GPT Assistants, then moves into multi-agent systems using CrewAI and AutoGen, and then into more sophisticated orchestration with behavior trees and platforms like Nexus. These chapters are technical, but not dry. Code examples are annotated clearly, and the tooling is open source. If you’re looking to build, not just read, you’ll find working projects on GitHub with guidance for local and API-hosted LLMs.
Lanham showcases a wide range of tools, but often presents them optimistically without much critique. There is little discussion of trade-offs, limitations, performance issues, or how these tools behave at scale or under real-world constraints.
The book gives lots of project examples, but they’re largely illustrative or experimental. It would benefit from at least one extended, real-world use case that shows how an agent system is deployed, maintained, and integrated into a broader business or operational environment.
Who is it for?
AI Agents in Action is not a book for nontechnical audiences or for those looking for abstract theory. It’s practical, applied, and sometimes dense. But that’s the point. Lanham is building a bridge between raw model capabilities and full agent systems. Keep in mind the book moves quickly and assumes a fair amount of comfort with Python, GitHub, and LLM APIs.
In short, if your goal is to go beyond talking about LLMs and start deploying intelligent, interactive systems, whether for internal workflows, customer-facing automation, or research prototypes, this book will save you a lot of trial and error. It doesn’t pretend to have all the answers, but it will enable you to ask better questions and build smarter agents.