Neo4j

What it is

The leading native graph database. Stores nodes, relationships and properties; queried with Cypher (a SQL-like graph language). Ships with the Graph Data Science library for centrality, community detection, similarity.

How Vaaani uses it

• Storing the knowledge graph behind a Graph RAG system
• Modeling org charts, supply chains, fraud rings
• Recommending products / content via graph traversal
• PageRank and Louvain community detection on enterprise data

Why it makes the cut

Graphs in Postgres are possible but painful. Neo4j makes 'find all 3-hop connections between X and Y' a one-liner.

Sample code

// Cypher
MATCH (p:Person {name: "Neil"})-[:WORKS_ON]->(proj)
MATCH (proj)<-[:USES]-(tool)
RETURN tool.name, count(*) AS uses
ORDER BY uses DESC;