I am writing a book: AI Agents with LangChain and LangGraph.
It is a practical developer book. 20 chapters in 5 parts, plus 6 complete projects:
- Part I, Foundations: what agents are, setup, prompts, structured output
- Part II, LangChain core: LCEL, chat models, tools, agents, memory
- Part III, RAG done properly: chunking, vector stores, retrievers, quality
- Part IV, LangGraph: state, routing, checkpointers, human in the loop, multi-agent systems
- Part V, Production: guardrails, cost, evals, deployment with FastAPI and Docker
One rule I set for myself while writing it: no code goes into the book unless it has been run first, and no output is printed unless it came from a real run. If a chapter's example fails, the chapter does not ship.
The book is also the in-depth companion to the AI Agent Engineering course I am launching on LogicStack. Enrollment opens 1 November and is free until 15 November. Students who enroll in that window and complete the course will receive the full book PDF free, together with their certificate.
The book page is already live, with the cover and the full chapter list:
https://logicstack.org/books/ai-agents-with-langchain-and-langgraph
If you want launch updates and the course outline PDF, join the WhatsApp group:
https://chat.whatsapp.com/G5tq6P9TATVHnZym5UgZNG
More chapter news as the writing continues.
