AI Agents and Applications: With LangChain, LangGraph, and MCP
by Manning (Roberto Infante)
A project-driven Python book that goes from your first programmatic prompt to tool-using LangGraph agents, advanced RAG and your own MCP servers.
Overview
AI Agents and Applications: With LangChain, LangGraph, and MCP is a 448-page Manning book by Roberto Infante, published in early 2026 (ISBN 9781633436541; ebook 9781638357827). It is organized in five parts. Part 1 introduces AI agents and applications and shows how to execute prompts programmatically. Part 2 covers summarization: summarizing text with LangChain, building a research summarization engine, and agentic workflows with LangGraph. Part 3 moves to Q&A chatbots, with RAG fundamentals on ChromaDB and chatbots built with LangChain and traced in LangSmith. Part 4 is three chapters of advanced RAG: advanced indexing, question transformations, and query generation, routing and retrieval postprocessing. Part 5 is about agents: tool-based agents in LangGraph, multi-agent systems, building and consuming MCP servers, and productionizing agents with memory and guardrails. Five appendixes cover trying out LangChain, setting up Jupyter, choosing an LLM, installing SQLite on Windows, and open-source LLMs. Manning's description also stresses context engineering, cost and latency tradeoffs, and evaluating and monitoring systems once they are in production. The companion GitHub repository (roberto-inf/building-llm-applications) has per-chapter folders ch01 to ch11 plus appendix E code. Infante builds agentic systems for quantitative analysis at a London-based hedge fund, which explains the book's focus on maintainable systems over demos.
At a Glance
- Topic
- Agentic
- Level
- Intermediate
- Format
- Book
- Cost
- Paid
- Duration
- 448-page book, 14 chapters plus 5 appendixes, self-paced with chapter-by-chapter Python code
- Provider
- Manning (Roberto Infante)
- Hands-on
- Yes — code/exercises
- Certificate
- None
What You’ll Learn
- ✓Execute prompts programmatically and structure LLM calls inside a Python application
- ✓Build a research summarization engine and turn it into an agentic LangGraph workflow
- ✓Implement RAG fundamentals with ChromaDB and build Q&A chatbots traced in LangSmith
- ✓Apply advanced indexing techniques to improve what a RAG pipeline can retrieve
- ✓Use question transformations, query generation and routing to raise retrieval quality
- ✓Build tool-based agents with LangGraph and coordinate several agents in a multi-agent system
- ✓Build your own MCP servers and consume existing ones from an agent
- ✓Productionize agents with memory, guardrails, evaluation and monitoring in mind
Highlights
- •Covers the current LangChain stack end to end in one book: LangChain, LangGraph, LangSmith and MCP, with a full chapter on building and consuming MCP servers
- •Three dedicated chapters on advanced RAG (indexing, question transformations, routing and postprocessing), which is more retrieval depth than most agent books carry
- •Companion GitHub repo with per-chapter code folders, roughly 129 stars and 94 forks when checked in November 2026
- •Written by a practitioner who builds agentic systems for quantitative finance, with an emphasis on cost, latency and maintainability
Who It’s For
Best For
- ✓Python developers moving from single LLM calls to multi-step agents and RAG applications
- ✓Engineers standardizing on LangChain and LangGraph who want a structured, book-length path
- ✓Teams that need to add MCP servers and production guardrails to an existing agent
Prerequisites
- •Working Python skills and comfort setting up a Jupyter environment
- •Access to an LLM API or a local open-source model (appendixes C and E help you choose)
- •No prior LangChain experience required; appendix A introduces it
FAQ
What is AI Agents and Applications: With LangChain, LangGraph, and MCP?
AI Agents and Applications is a 2026 Manning book by Roberto Infante for Python developers who want to build production LLM applications with LangChain, LangGraph and LangSmith. It works through summarization engines, RAG chatbots, advanced retrieval, tool-based and multi-agent systems, and MCP servers, so by the end you can design, build and harden an agent yourself.
Is AI Agents and Applications: With LangChain, LangGraph, and MCP free?
AI Agents and Applications: With LangChain, LangGraph, and MCP is a paid resource.
What level is AI Agents and Applications: With LangChain, LangGraph, and MCP for?
AI Agents and Applications: With LangChain, LangGraph, and MCP is aimed at a intermediate audience. Recommended background: Working Python skills and comfort setting up a Jupyter environment, Access to an LLM API or a local open-source model (appendixes C and E help you choose), No prior LangChain experience required; appendix A introduces it.
How long does AI Agents and Applications: With LangChain, LangGraph, and MCP take?
Expect roughly 448-page book, 14 chapters plus 5 appendixes, self-paced with chapter-by-chapter Python code. Most learners work through it at their own pace.
What will I learn from AI Agents and Applications: With LangChain, LangGraph, and MCP?
You'll learn: Execute prompts programmatically and structure LLM calls inside a Python application; Build a research summarization engine and turn it into an agentic LangGraph workflow; Implement RAG fundamentals with ChromaDB and build Q&A chatbots traced in LangSmith; Apply advanced indexing techniques to improve what a RAG pipeline can retrieve; Use question transformations, query generation and routing to raise retrieval quality; Build tool-based agents with LangGraph and coordinate several agents in a multi-agent system; Build your own MCP servers and consume existing ones from an agent; Productionize agents with memory, guardrails, evaluation and monitoring in mind.
Topics
Sources
This page was written from 3 sources, 2 on domains other than manning.com.