Learning path

Generative AI & GraphRAG

Give your LLMs a knowledge graph memory. Build RAG pipelines that retrieve structured, contextual facts for more accurate, grounded answers.

Large language models are fluent but forgetful: they hallucinate, lose track of how things connect, and cannot cite where an answer came from. A knowledge graph gives them a memory, and GraphRAG — retrieval that combines vector search over unstructured text with graph traversal over structured relationships — lets them answer from grounded, traceable data. Begin with Neo4j & GenerativeAI Fundamentals for the concepts every later course builds on, or get something working first with Building Agents in Neo4j Aura.

Browse courses
  • 10
    Courses
  • ~18h
    Total time

The curriculum

10 courses, to complete.

A rhythm of deeper courses that teach concepts, and quicker labs that drill a single pattern.

10 Courses · 1–2 hours each
  1. 01
    Course2h

    Neo4j & GenerativeAI Fundamentals

    Start here for the core concepts — vectors, embeddings, and how a graph grounds an LLM.

    Build a Neo4j-backed ApplicationContext EngineerDevelopment
    Start the course
  2. 02
    Course2h

    Building Agents in Neo4j Aura

    Build and run your first graph-backed agent on Aura before going deeper into the plumbing.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  3. 03
    Course1h

    Introduction to Vector Indexes and Unstructured Data

    Store and search embeddings directly in Neo4j to retrieve over unstructured text.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  4. 04
    Course1h

    Building Knowledge Graphs with LLMs

    Turn raw documents into a structured knowledge graph using LLMs.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  5. 05
    Course2h

    Constructing Knowledge Graphs with Neo4j GraphRAG for Python

    Assemble full GraphRAG pipelines in Python with the neo4j-graphrag package.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  6. 06
    Course1h

    Using Neo4j with LangChain

    Wire your graph into LangChain applications, chains, and retrievers.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  7. 07
    Course2h 30m

    Context Graphs: Agent Memory with Neo4j

    Give agents durable, queryable memory with a context graph.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  8. 08
    Course2h

    Developing with Neo4j MCP Tools

    Expose your graph to AI agents through the Model Context Protocol.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  9. 09
    Course2h

    Building GraphRAG Python MCP tools

    Build your own GraphRAG MCP tools in Python.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course
  10. 10
    Course2h

    Building GraphRAG TypeScript MCP tools

    The same custom-tool build, in TypeScript, for JS/TS stacks.

    Build a Neo4j-backed ApplicationContext EngineerGenerative AI & GraphRAG
    Start the course

Building a startup?

Apply today to the Neo4j Startup Program for Aura credits and technical guidance sessions.

Apply now