Learning LangChain: Building AI and LLM Applications with LangChain and LangGraph

Preface

Brief Primer on LLMs

Instruction-Tuned LLMs

Dialogue-Tuned LLMs

Fine-Tuned LLMs

Brief Primer on Prompting

Zero-Shot Prompting

Chain-of-Thought

Retrieval-Augmented Generation

Tool Calling

Few-Shot Prompting

LangChain and Why It’s Important

What to Expect from This Book

Conventions Used in This Book

Using Code Examples

O’Reilly Online Learning

How to Contact Us

Acknowledgments

1. LLM Fundamentals with LangChain

Getting Set Up with LangChain

Using LLMs in LangChain

Making LLM Prompts Reusable

Getting Specific Formats out of LLMs

JSON Output

Other Machine-Readable Formats with Output Parsers

Assembling the Many Pieces of an LLM Application

Using the Runnable Interface

Imperative Composition

Declarative Composition

Summary

2. RAG Part I: Indexing Your Data

The Goal: Picking Relevant Context for LLMs

Embeddings: Converting Text to Numbers

Embeddings Before LLMs

LLM-Based Embeddings

Semantic Embeddings Explained

Converting Your Documents into Text

Splitting Your Text into Chunks

Generating Text Embeddings

Storing Embeddings in a Vector Store

Getting Set Up with PGVector

Working with Vector Stores

Tracking Changes to Your Documents

Indexing Optimization

MultiVectorRetriever

RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval

ColBERT: Optimizing Embeddings

Summary

3. RAG Part II: Chatting with Your Data

Introducing Retrieval-Augmented Generation

Retrieving Relevant Documents

Generating LLM Predictions Using Relevant Documents

Query Transformation

Rewrite-Retrieve-Read

Multi-Query Retrieval

RAG-Fusion

Hypothetical Document Embeddings

Query Routing

Logical Routing

Semantic Routing

Query Construction

Text-to-Metadata Filter

Text-to-SQL

Summary

4. Using LangGraph to Add Memory to Your Chatbot

Building a Chatbot Memory System

Introducing LangGraph

Creating a StateGraph

Adding Memory to StateGraph

Modifying Chat History

Trimming Messages

Filtering Messages

Merging Consecutive Messages

Summary

5. Cognitive Architectures with LangGraph

Architecture #1: LLM Call

Architecture #2: Chain

Architecture #3: Router

Summary

6. Agent Architecture

The Plan-Do Loop

Building a LangGraph Agent

Always Calling a Tool First

Dealing with Many Tools

Summary

7. Agents II

Reflection

Subgraphs in LangGraph

Calling a Subgraph Directly

Calling a Subgraph with a Function

Multi-Agent Architectures

Supervisor Architecture

Summary

8. Patterns to Make the Most of LLMs

Structured Output

Intermediate Output

Streaming LLM Output Token-by-Token

Human-in-the-Loop Modalities

Multitasking LLMs

Summary

9. Deployment: Launching Your AI Application into Production

Prerequisites

Install Dependencies

Large Language Model

Vector Store

Backend API

Create a LangSmith Account

Understanding the LangGraph Platform API

Data Models

Features

Deploying Your AI Application on LangGraph Platform

Create a LangGraph API Config

Test Your LangGraph App Locally

Deploy from the LangSmith UI

Launch LangGraph Studio

Security

Summary

10. Testing: Evaluation, Monitoring, and Continuous Improvement

Testing Techniques Across the LLM App Development Cycle

The Design Stage: Self-Corrective RAG

The Preproduction Stage

Creating Datasets

Defining Your Evaluation Criteria

Regression Testing

Evaluating an Agent’s End-to-End Performance

Production

Tracing

Collect Feedback in Production

Classification and Tagging

Monitoring and Fixing Errors

Summary

11. Building with LLMs

Interactive Chatbots

Collaborative Editing with LLMs

Ambient Computing

Summary

Index

About the Authors

If you're looking to build production-ready AI applications that can reason and retrieve external data for context-awareness, you'll need to master--;a popular development framework and platform for building, running, and managing agentic applications. LangChain is used by several leading companies, including Zapier, Replit, Databricks, and many more. This guide is an indispensable resource for developers who understand Python or JavaScript but are beginners eager to harness the power of AI.

Authors Mayo Oshin and Nuno Campos demystify the use of LangChain through practical insights and in-depth tutorials. Starting with basic concepts, this book shows you step-by-step how to build a production-ready AI agent that uses your data.

• Harness the power of retrieval-augmented generation (RAG) to enhance the accuracy of LLMs using external up-to-date data
• Develop and deploy AI applications that interact intelligently and contextually with users
• Make use of the powerful agent architecture with LangGraph
• Integrate and manage third-party APIs and tools to extend the functionality of your AI applications
• Monitor, test, and evaluate your AI applications to improve performance
• Understand the foundations of LLM app development and how they can be used with LangChain