Glossary
What is Knowledge Graph?
A knowledge graph is a data structure that represents information as typed nodes connected by typed relationships. Each node represents an entity (a concept, person, document, or decision) and each edge represents a relationship between entities. Knowledge graphs support traversal queries, structured reasoning, and source attribution that flat document stores cannot.
Last updated:
Capabilities
What Knowledge Graph does
Stores typed entities
Each node has a type (concept, pattern, person, decision, customer) with structured metadata like status, confidence, and source.
Stores typed relationships
Edges between nodes carry meaning. Not just 'X links to Y' but 'X depends on Y' or 'X contradicts Y' or 'X is evidence for Y'.
Supports traversal queries
Queries can follow relationships across multiple hops: 'show me all customer evidence linked to ICP X via positioning Y'.
Powers AI retrieval
AI tools can search the graph for semantically related nodes and traverse relationships to gather grounded context.
Distinctions
Knowledge Graph vs adjacent concepts
Knowledge Graph is often confused with related but distinct ideas. Here is how it differs.
| Concept | What it is | How Knowledge Graph differs |
|---|---|---|
| Context OS | The data structure: typed nodes and typed relationships. | The system around the graph: ingestion, query interface, MCP server, role-based access, AI integration. The graph is one component of a Context OS. |
| Relational database | Tables, rows, columns, foreign keys. Optimized for transactional queries. | Graph traversal across many entity types and relationship types. Optimized for connected reasoning over heterogeneous data. |
| Vector database | Stores embeddings for semantic similarity search. | Stores typed structure with status, confidence, and explicit relationships. Often used together with vector search, not as a replacement. |
Who uses it
Who uses Knowledge Graph
Search engines (Google's Knowledge Graph), AI retrieval systems, enterprise search platforms, and Context OS implementations like DearTech-OS use knowledge graphs to support structured reasoning over connected information.
FAQ
Common questions about Knowledge Graph
Is a Context OS a knowledge graph?
A Context OS uses a knowledge graph as a core component, but it is more than the graph itself. A Context OS includes ingestion workflows, a query interface (typically MCP or API), role-based access, source attribution, and AI tool integration that surround the graph.
What kinds of queries does a knowledge graph support?
Traversal queries that follow typed relationships across many hops, neighborhood queries that find all nodes connected to a starting node, type-filtered searches that look only at certain entity types, and structured queries that combine graph traversal with metadata filters like status or confidence.
How is a knowledge graph different from a vector database?
A vector database stores embeddings for semantic similarity. A knowledge graph stores typed structure for navigable relationships. They are often used together: vector search finds candidate nodes, then graph traversal walks the relationships.
Related terms
Keep going
See Knowledge Graph in practice
DearTech-OS is a Context OS for founder-operators. Explore the product or talk through whether one is right for your team.