Leading With Data· a Bicycle Guide

Partial book profile

Prompt Engineering for LLMs The Art and Science of Building Large Language Model-Based Applications

A practical guide for software engineers on building reliable and powerful applications using Large Language Models by mastering the art and science of prompt engineering.

The full book profile is not available yet.

Get the book →

What it’s about

Written by two key engineers behind GitHub Copilot, one of history's most successful commercial AI products, "Prompt Engineering for LLMs" is the definitive guide for application developers aiming to harness the revolutionary power of Large Language Models. This book demystifies LLMs, revealing their core function as text completion engines and teaching you how to communicate with them effectively. It provides a comprehensive framework for building LLM-powered applications, covering everything from foundational principles and core prompting techniques like few-shot learning and Retrieval-Augmented Generation (RAG), to advanced strategies for creating conversational agents and complex automated workflows. By mastering the systematic, engineering-focused approaches within, you will gain the confidence and skills to build the next generation of robust, valuable, and user-friendly AI-powered software.

The through-line

Who it’s for
Application engineers and software developers who want to build the next generation of software by harnessing the revolutionary power of Large Language Models. They aim to create robust, reliable, and valuable LLM-powered applications, moving beyond simple demos to production-ready products.
The problem
Building high-quality applications on top of LLMs is difficult; the models are often unpredictable, prone to hallucination, and it's not obvious how to reliably translate a user's problem into a format the LLM can consistently and correctly solve. Developers feel uncertain and overwhelmed by the rapidly changing paradigm of LLM development. They are frustrated when their efforts result in inconsistent or incorrect outputs, making them doubt their ability to build production-ready AI products.
The plan
  1. Understand the Foundations: Learn precisely how LLMs work, from tokenization and transformers to the evolution from text completion to chat and tool use.
  2. Master Core Prompting Techniques: Learn to source content, assemble effective prompts using methods like few-shot learning and RAG, and tame the model's output.
  3. Build Advanced Applications: Create sophisticated conversational agents, design complex multi-step LLM workflows, and systematically evaluate your application's quality.
The payoff
You are equipped with the knowledge and confidence to build the next generation of powerful, reliable LLM-powered applications. · You can reliably transform ideas into a model-friendly format, creating high-quality, valuable products for users. · You become an expert 'LLM wrangler,' skillfully guiding AI to solve complex problems and unlock new potential in any domain.

See our guide

Related profiles we’ve built

Featured in these guides

Co-cited books