Название: Prompt Engineering in Action (MEAP v4)
Автор: Naresh Vurukonda, Shivendra Srivastava
Издательство: Manning Publications
Год: 2025
Страниц: 244
Язык: английский
Формат: pdf (true), epub
Размер: 21.1 MB

Tested techniques for writing excellent AI prompts!

Each LLM seems to have a mind of its own, and it can be challenging to get the exact results you want. Prompt Engineering in Action teaches you to write prompts that generate the text and code you want, regardless of the LLM you choose. In it, you’ll discover structured and reusable prompt patterns that reduce hallucinations, customize LLMs to specific tasks, and improve the quality of your code generation.

Prompt Engineering in Action teaches you practical prompt engineering skills including:

Designing context-aware prompts tailored for specific tasks
Understanding and minimizing hallucinations
When to use prompt patterns such as Persona, Recipe, Template, and Game-Play
Utilizing templates to ensure consistent and reusable outputs
Integrating external knowledge bases with Retrieval-Augmented Generation (RAG)
Building and deploying practical LLM-based apps using LangChain

Prompt Engineering in Action presents patterns, templates, and techniques that help you get consistent, valuable responses from LLMs. You’ll learn how to design precise and context-aware prompts, discover metrics you can use to assess prompt quality, learn methods to scale and collaborate on prompts, and build advanced and agentic AI apps using LangChain.

about the book:
Prompt Engineering in Action teaches universal prompting principles, each illustrated with concise examples of a prompt in action. You’ll find more than just simple prompt-and-response methods. It goes beyond the basics of prompting to explore reusability, scaling for prompts, and managing stochastic responses. Dive into prompt patterns such as Persona, Game-Play, and Recipe patterns, and into emerging techniques such as Retrieval-Augmented Generation (RAG) and dynamic agents. You’ll soon be building solutions that are efficient, accurate, and highly adaptable for real-world applications.

about the reader:
For readers familiar with the basics of working with LLMs. No specialist knowledge required.