Notes from the graph & GenAI frontier.
Tutorials, patterns and field notes on building GraphRAG systems, knowledge graphs and LLM agents on Neo4j — written by the people who build them.
Featured
Editor's Picks
Traversing Relationships in the 2026 World Cup
Traverse the full 2026 World Cup in Neo4j to build group tables, a golden-boot ranking, and the champions' path, using multi-hop Cypher, WITH pipelines, and aggregation.
Loading 2026 World Cup Data with Cypher
Load all 104 matches and every goal of the 2026 World Cup from open data into AuraDB, using constraints and idempotent Cypher you can re-run confidently.
Introducing the New GraphAcademy
Our first full redesign since 2024 — with teams, a revamped Neo4j Fundamentals course, editable data models, and a new onboarding assistant.
Browse Posts
Find your next read
RAG vs GraphRAG - Two Retrievers, One Question
Same movie question, two retrievers, two answers. See how GraphRAG beats plain vector RAG on Neo4j's recommendations dataset, then learn to build it.
How Cypher Graph Types Kept Our AI-Generated Queries Honest During a Migration
We rebuilt GraphAcademy against the same Neo4j database. Graph types made sure AI-generated Cypher could not drift from the data model.