Why Vectors Aren’t Enough
Engine Room Article 8: The Case for Structured Knowledge
The Vector Search Pattern
The current approach to grounding AI in organizational knowledge typically involves vector databases: embed your documents, do similarity search at query time, feed relevant chunks to the language model.
Where it can struggle: when the answer requires connecting concepts that aren’t semantically similar, when relationships matter more than content similarity.
Vector search finds semantically similar content. But expertise often involves connecting concepts that aren’t similar - they’re related in ways similarity search can miss.
What Graphs Add
Knowledge graphs make relationships first-class objects. Instead of inferring connections from attribute matching, you explicitly represent how things relate.
Building a knowledge graph from nutritional and compound databases taught me how much value lives in relationships rather than entities. That experience has shaped how I think about current approaches to RAG and retrieval.
Graphs make relationships first-class objects. The valuable questions are often about traversing connections, not finding similar content.
Complementary Approaches
This isn’t an argument that graphs replace vectors - they’re complementary. Vector search excels at finding relevant content when you’re not sure exactly what you’re looking for. Graph traversal excels when you know the relationship structure and need to reason through connections.
Vectors and graphs solve different problems. Vectors find similar content; graphs traverse relationships. Complex knowledge applications often benefit from both.
Related: 07-source—engine-room-series