LocalGraph Tutorials

Welcome to AI Agents 101! Master production-ready AI agents that run entirely on your machine using LangGraph and Ollama.

What You'll Build

Through 25 hands-on tutorials, you'll master the patterns and techniques used in production AI agents:

Core Patterns: Build conversational agents with tools, memory, and human oversight
RAG Systems: Create search and research assistants with document retrieval
Multi-Agent Systems: Coordinate specialized agents for complex workflows
Advanced Reasoning: Implement cutting-edge patterns like Reflexion and LATS

Each tutorial includes:

Step-by-step implementation guide
Working code examples
Interactive playground to test your agent
Quiz to verify understanding
Progress tracking across your learning journey

Getting Started

New to LangGraph? Start here:

Setup Guide - Install dependencies and verify your environment
Tutorial 01: Chatbot Basics - Learn LangGraph fundamentals
Work through tutorials sequentially or jump to patterns that interest you

Phase 1: Core Patterns

Core Patterns

0/7 completed

Master the fundamental patterns that power all LangGraph applications. These tutorials teach you how to build conversational agents with tool calling, memory, human oversight, and self-improvement capabilities.

Tutorial 01

Chatbot Basics

Learn StateGraph, nodes, edges, and message handling. Build your first conversational agent with streaming support.

Core Patterns

Tutorial 02

Tool Calling

Implement the ReAct pattern from scratch. Give your agent access to tools and let it decide when to use them.

Core Patterns

Tutorial 03

Memory & Persistence

Add conversation memory with checkpointers. Implement thread-based conversations that persist across sessions.

Core Patterns

Tutorial 04

Human-in-the-Loop

Pause execution for human approval. Implement approval workflows and resume from saved checkpoints.

Core Patterns

Tutorial 05

Reflection

Build generate-critique-revise loops. Teach your agent to improve its own outputs through self-criticism.

Core Patterns

Tutorial 06

Plan & Execute

Separate planning from execution. Use structured outputs to build multi-step workflows with replanning.

Core Patterns

Tutorial 07

Research Assistant

Capstone project combining all core patterns. Build a research assistant with web search and citation tracking.

Core Patterns

Phase 2: RAG Patterns

RAG Patterns

0/6 completed

Learn to build retrieval-augmented generation systems that ground LLM responses in your documents. Progress from basic retrieval to advanced patterns with quality grading, web search fallbacks, and agentic control.

Tutorial 08

Basic RAG

Document loading, chunking, embeddings, and ChromaDB. Build a simple question-answering system.

RAG Patterns

Tutorial 09

Self-RAG

Add document grading and hallucination detection. Implement retry loops for better answer quality.

RAG Patterns

Tutorial 10

CRAG (Corrective RAG)

Web search fallback when documents aren't relevant. Combine local retrieval with external knowledge.

RAG Patterns

Tutorial 11

Adaptive RAG

Query classification and strategy routing. Choose between retrieval, web search, or direct generation.

RAG Patterns

Tutorial 12

Agentic RAG

Agent-controlled retrieval with multi-step reasoning. Let the agent decide when and what to retrieve.

RAG Patterns

Tutorial 13

Perplexity Clone

Build a research assistant with citations and source metadata. Implement follow-up question flows.

RAG Patterns

Phase 3: Multi-Agent Patterns

Multi-Agent Patterns

0/7 completed

Coordinate multiple specialized agents to tackle complex tasks. Learn patterns for collaboration, hierarchies, handoffs, and evaluation of multi-agent systems.

Tutorial 14

Multi-Agent Collaboration

Supervisor coordinates researcher, coder, and reviewer agents. Learn agent orchestration patterns.

Multi-Agent

Tutorial 15

Hierarchical Teams

Nested team structures with managers and workers. Build scalable multi-agent hierarchies.

Multi-Agent

Tutorial 16

Subgraphs

Composable graph components for reusable agent teams. Encapsulate complex workflows.

Multi-Agent

Tutorial 17

Agent Handoffs

Seamless transitions between specialized agents. Implement context-preserving handoff patterns.

Multi-Agent

Tutorial 18

Agent Swarm

Decentralized agent coordination without supervisors. Emergent behavior from agent interactions.

Multi-Agent

Tutorial 19

Map-Reduce Agents

Parallel processing with result aggregation. Scale work across multiple agents efficiently.

Multi-Agent

Tutorial 20

Multi-Agent Evaluation

Testing and debugging multi-agent systems. Metrics, logging, and quality assurance strategies.

Multi-Agent

Phase 4: Advanced Reasoning

Advanced Reasoning

0/5 completed

Implement cutting-edge research patterns for complex reasoning tasks. These advanced techniques push the boundaries of what local LLMs can accomplish.

Tutorial 21