Context Graph

A context graph is a structured map of how a company’s information connects: which documents, decisions, people and conversations relate to each other, and how.

A search index treats each document as an isolated result to be ranked. A graph treats the relationships as the data. It records that a contract belongs to an account, that a decision in a thread superseded an earlier one, that a person owns a workstream, and that a meeting produced the note now being cited. Those edges are what let a system answer questions no single document contains.

The distinction becomes concrete when something is out of date. A ranked list will happily return a superseded document, because it is still the best keyword match. A graph knows something replaced it, and can say so, which is the difference between retrieving text and understanding a company.

Frequently Asked Questions

A knowledge graph typically models entities and facts about the world or a domain. A context graph models a specific company’s working state: its people, records, decisions and the connections between them.