Tutorials

Learn neo4j-agent-memory by building real applications. These tutorials guide you through complete, working examples from start to finish.

Tutorials are learning-oriented. They take you by the hand through a series of steps to complete a project. They focus on building a context graph—the personalized knowledge structure that makes AI agents truly intelligent.

Getting Started

Pick the path that matches your stack:

Tutorial	What You’ll Build	Time
NAMS Quickstart (Python)	The hosted-service path. Get an API key, install the SDK, run your first memory operations — no Neo4j to operate. Ideal for prototypes and SaaS deployments.	~10 min
Hosted Quickstart (TypeScript)	Same hosted-service path, TypeScript flavor. Works on Node, Bun, Deno, Cloudflare Workers, and Vercel Edge.	~10 min
Build Your First Memory-Enabled Agent (Self-Hosted)	A working memory system against your own Neo4j. Stores conversations, extracts entities automatically, enables semantic search. You’ll see your context graph in Neo4j Browser.	~30 min
First Agent Memory (TypeScript, hosted)	TypeScript end-to-end: NAMS + Vercel AI SDK middleware, multi-turn chat with three-tier context recall.	~20 min

Tutorial

What You’ll Build

Time

NAMS Quickstart (Python)

The hosted-service path. Get an API key, install the SDK, run your first memory operations — no Neo4j to operate. Ideal for prototypes and SaaS deployments.

~10 min

Hosted Quickstart (TypeScript)

Same hosted-service path, TypeScript flavor. Works on Node, Bun, Deno, Cloudflare Workers, and Vercel Edge.

~10 min

Build Your First Memory-Enabled Agent (Self-Hosted)

A working memory system against your own Neo4j. Stores conversations, extracts entities automatically, enables semantic search. You’ll see your context graph in Neo4j Browser.

~30 min

First Agent Memory (TypeScript, hosted)

TypeScript end-to-end: NAMS + Vercel AI SDK middleware, multi-turn chat with three-tier context recall.

~20 min

Building Real Applications

After completing the first tutorial, build complete applications:

Tutorial	What You’ll Build	Time
Add Conversation Memory to a Chatbot	A shopping assistant chatbot that remembers users, learns preferences (brands, sizes, budget), and provides personalized recommendations across sessions.	~45 min
Build a Knowledge Graph from Documents	Process financial documents to extract entities (companies, people, securities), discover relationships, and power intelligent Q&A with LLM integration.	~45 min
Build a Strands Agent with Context Graph	Create an AI agent using AWS Strands SDK with persistent memory backed by Neo4j. Uses Amazon Bedrock for LLM and embeddings.	~30 min
Connect Claude Desktop to Your Knowledge Graph	Set up the MCP server so Claude Desktop can store conversations, extract entities, and query your knowledge graph automatically.	~15 min

Tutorial

What You’ll Build

Time

Add Conversation Memory to a Chatbot

A shopping assistant chatbot that remembers users, learns preferences (brands, sizes, budget), and provides personalized recommendations across sessions.

~45 min

Build a Knowledge Graph from Documents

Process financial documents to extract entities (companies, people, securities), discover relationships, and power intelligent Q&A with LLM integration.

~45 min

Build a Strands Agent with Context Graph

Create an AI agent using AWS Strands SDK with persistent memory backed by Neo4j. Uses Amazon Bedrock for LLM and embeddings.

~30 min

Connect Claude Desktop to Your Knowledge Graph

Set up the MCP server so Claude Desktop can store conversations, extract entities, and query your knowledge graph automatically.

~15 min

What You’ll Learn

These tutorials cover the core concepts of neo4j-agent-memory:

Context Graph: Build and query the knowledge structure that powers personalization
Entity Extraction: Automatically extract people, companies, products from text
Preference Learning: Capture and use user preferences for personalization
Semantic Search: Find information by meaning, not just keywords
Multi-Session Persistence: Memory that survives across conversations
Agent Integration: Connect memory to LLMs for intelligent responses

Enterprise Examples

Each tutorial includes examples for common enterprise use cases:

Financial Services: Client profiles, investment preferences, document analysis
Ecommerce Retail: Customer preferences, product recommendations, purchase history

Before You Start

The Python tutorials assume:

Python 3.10 or later installed
Basic familiarity with Python and async programming
Either an API key from NAMS (hosted path) or access to a Neo4j database (self-hosted path; setup instructions included)

The TypeScript tutorials assume:

Node.js 20 or later
An API key from NAMS
Basic familiarity with TypeScript and async/await

Tutorial Philosophy

These tutorials follow the Diataxis "learning-oriented" approach:

Step-by-step: Every action is explicit, with exact commands to run
Hands-on: You’ll build something real, not just read about concepts
One path: No choices or alternatives - just follow along
Complete: Each tutorial stands alone and produces a working result

If you’re looking for quick answers to specific problems, see the How-To Guides.

If you want to understand the concepts behind the library, see the Explanation section.

If you need to look up configuration options or API details, see the Reference section.

Getting Help

If you get stuck:

Check the Configuration Reference for environment setup
See the GitHub Discussions for community help
Report issues at GitHub Issues