There are, by now, quite a lot of LLM-related books to choose from, and I'm glad I got this one. It gives you a great overview of and deep dive into the necessary steps to create applications incorporating LLMs, including how to construct effective prompts and all the details that go into that, as well as how to "think like an LLM" to make use of them effectively. It also discusses tool usage and RAG, agents, as well as the all-important evals to make sure your application works, and keeps working, as expected.

It's filled with small nuggets of wisdom and insightful comments that can only come from people who've been actively applying their knowledge for years already, and I'm happy to be able to use that to get a jump-start for applying it in my own work.

Prompt Engineering for LLMs is a well-crafted introduction to one of the most important emerging skills in the age of AI: communicating effectively with large language models. Whether you're building LLM-powered apps or just trying to get better responses from AI, this book will help you become an LLM Whisperer.

Authors John Berryman and Albert Ziegler strike a great balance—delivering a thorough, yet not overly academic, treatment of key concepts in prompt engineering. I especially appreciated the way they frame both the philosophical underpinnings and the practical techniques. They provide a solid overview of LLM history and architecture, then dive into effective prompting strategies like few-shot learning, chain-of-thought prompting, and RAG.

The illustrations are clean and thoughtfully done—they complement the text rather than distract from it. And the book is refreshingly light on code, which makes it easier to focus on understanding the why behind prompt engineering instead of just the how.

Highly recommended for developers, tech leads, or anyone serious about tapping into the full potential of modern AI.

I just finished reading Prompt Engineering for LLMs. The authors have done a fantastic job. I've recommended the book to a few people already. Even though my team has worked on a fair amount of LLM implementations the past two years, I found something insightful in every chapter: either some new strategy to employ or a clarifying thought about some practice I had stumbled into in prompt engineering and previously had difficulty explaining to others.

My dev team has weekly internal talks led by team members on various topics they're interested in, and I've been wanting to do more with them around AI best practices. This book is giving me a good blueprint to hopefully start that. So thank you!

Many things to think about, to do, conceptually, but almost nothing on how to actually do it, if you are looking for that

Excellent information which I will be able to use.

Excellent resource! This book was incredibly helpful in growing my ability to articulate and improve my understanding of state of the art prompt engineering. I’ve recommended this book to several others and have heard similar positive reviews. Many thanks to the authors!

For how much this book costs, there's not enough about actual prompt engineering in this book to justify the cost. I thought the book would cover more advanced information about building prompts, types of prompts, how to structure prompts for different types of output.

This is purportedly an advanced book about prompt engineering. There's no need for a history of LLMs. There are a multitude of books that do that already. There's only 59 pages out of close to 300 that cover actual prompting. This was a disppontment.

Prompt Engineering for LLMs: The Art and Science of Building Large Language Model-Based Applications is a practical, insightful guide that meets software engineers exactly where they are; at the crossroads of traditional development and the fast-evolving world of AI integration. As large language models become embedded in both customer-facing products and internal systems, the ability to work fluently with them is quickly becoming a fundamental engineering skill. This book makes the case for that future and it provides a path to get there.

John Berryman and Albert Ziegler succeed not just in defining what prompt engineering is, but in teaching how to do it well. They present prompt engineering as both a science and an art, rooted in clear communication, structured thinking, and a deep understanding of how LLMs interpret and respond to language. The book moves fluently from foundational concepts to advanced techniques like retrieval-augmented generation, looped inference workflows, and conversational agency design. Whether you’re optimizing prompts for performance or architecting entire LLM-based systems, the material is grounded, accessible, and deeply applicable.

Perhaps most importantly, this book reflects a truth too often overlooked: prompt engineering isn’t a fringe specialization / job (as is being shown by that job currently going out of fashion at the time of writing). It’s becoming a core competency in software development. If you’re already building applications, or expect to in a world increasingly shaped by AI, this is a book you should take a strong look at.

An insightful and comprehensive introduction to building applications using foundation models.

I was a bit skeptical of this book when ordering, but it really impressed me. It doesn't give many concrete examples, but rather talks about how prompts work in relation to an AI's training data and how you can leverage knowledge you have of that to get the best performance from an LLM. There's lots of scenarios and lots of contingencies. Has helped me a lot when writing props for agents.

Comprehensive and very clear. You will end up with deep knowledge on the topic - and very much more efficient at your day job.

I neither liked the book nor disliked it. I didn't like that there weren't enough real examples and comparisons of prompts.

This book manages to include enough technical detail and principles to make it interesting and usable for developers, while also distilling more complex topics for all tech enthusiasts to understand and enjoy.

I particularly enjoyed the tips throughout the book and diagrams that I will surely revisit.

To list some of the concepts covered in this book:

- Transformer architecture / LLM history

- Little Red Hiding Hood principle

- Zero-shot, few-shot and chain-of-thought prompting

- Elastic snippets

- logprobs

- ReAct

- Tools

- Offline and online evaluation

And much, much more.

Highly recommended!