This source is a book titled "Prompt Engineering for LLMs: The Art and Science of Building Large Language Model-Based Applications" by John Berryman and Albert Ziegler, early engineers for Github Copilot, intended to guide readers in developing applications powered by large language models. It covers foundational concepts of prompt engineering, such as how LLMs work, and moves into core techniques including structuring prompt content, assembling prompts effectively, and taming the model's output. Furthermore, the book addresses advanced techniques like conversational agency and LLM-driven workflows and ends by presenting techniques for evaluating LLMs. The overall goal is to equip readers with the knowledge and skills to build complete LLM-based applications rather than just tweaking individual prompts for single, optimal outputs.

Ver todos los episodios