# What is a knowledge graph?

> A knowledge graph stores facts as things and the links between them, like client, contract, renewal date. AI can use one to answer connected questions.

Source: https://learn.ynteractive.com/content/glossary/knowledge-graph · Updated 2026-10-20 · Free from Stepthrough (https://learn.ynteractive.com)

[AI agents](https://learn.ynteractive.com/content/glossary#topic-ai-agents) · [AI glossary](https://learn.ynteractive.com/content/glossary)

Quick answer

A knowledge graph stores information as things, such as clients, contracts and people, and the named links between them, such as "signed", "manages" or "renews on". Instead of searching text, you can follow the links: which clients does Dana manage whose contracts renew this quarter?

Last updated October 20, 2026

[Take the free module →](https://learn.ynteractive.com/courses/ai-for-law-firms/10)Module 10 of the free [AI for Law Firms](https://learn.ynteractive.com/courses/ai-for-law-firms) course, about 7 minutes. No account needed.

Updated October 20, 2026 · By [Robert Breen](https://learn.ynteractive.com/content/about)

## Why it matters for a small business

Google has used a knowledge graph since 2012 to show facts in search results. In AI, knowledge graphs are paired with language models in approaches like Microsoft's GraphRAG, which builds a graph from documents to answer questions that span many of them.

Most small teams don't need to build one. Your CRM and spreadsheets already hold linked records. The idea is useful because it explains why "find the passage" tools struggle with "connect the dots across ten files" questions.

Some AI products now say they use a graph behind the scenes to link people, files and meetings, for example to answer who worked on a project. When a tool claims that, the same checks apply: ask where each link came from and test it on questions you can verify.

## In a real lesson: NotebookLM (Now Gemini Notebook): Cited Answers and a Briefing from Your Docs

In [NotebookLM (Now Gemini Notebook)](https://learn.ynteractive.com/content/notebooklm-gemini-notebook-cited-answers-briefing-doc), Jess at Harbor Light Cleaning asks one question across three sources: the Northgate Dental cleaning agreement, her crew handbook and the services page. The answer finds a conflict: the contract requires a hospital-grade disinfectant, while the crew kit lists only a green cleaner.

The notebook finds that by reading the sources, not by keeping a knowledge graph. Its Studio panel can also make a Mind Map, which looks similar but is a visual summary of topics, not a database of linked facts you can query.

[Try this lesson free](https://learn.ynteractive.com/modules/gemini-notebook) or [read the step-by-step guide](https://learn.ynteractive.com/content/notebooklm-gemini-notebook-cited-answers-briefing-doc).

## Common confusions

### Knowledge graph vs mind map

A mind map is a picture of topics for people to read. A knowledge graph is structured data a system can query.

### Knowledge graph vs knowledge base

A knowledge base is a set of documents. A knowledge graph is facts and relationships stored as data.

## Tips

- For connected questions, start with the systems that already link records, like your CRM.
- Ask AI to list the source for each connection it claims.

## Related terms

[Knowledge base](https://learn.ynteractive.com/content/glossary/knowledge-base) · [Retrieval-augmented generation (RAG)](https://learn.ynteractive.com/content/glossary/rag) · [Vector database](https://learn.ynteractive.com/content/glossary/vector-database) · [Semantic search](https://learn.ynteractive.com/content/glossary/semantic-search) · [CRM](https://learn.ynteractive.com/content/glossary/crm)

## More AI agents terms

[LangChain](https://learn.ynteractive.com/content/glossary/langchain) · [LLM observability](https://learn.ynteractive.com/content/glossary/llm-observability) · [Make AI Agents](https://learn.ynteractive.com/content/glossary/make-ai-agents) · [Max iterations (agent step limit)](https://learn.ynteractive.com/content/glossary/max-iterations)

## Where you use it: free lessons

- [NotebookLM (Now Gemini Notebook): Cited Answers and a Briefing from Your Docs](https://learn.ynteractive.com/content/notebooklm-gemini-notebook-cited-answers-briefing-doc) (Gemini Notebook, 7 min)
- [Answer Staff Questions from Your Own Docs using n8n + Pinecone](https://learn.ynteractive.com/content/n8n-ai-agent-pinecone-rag-chatbot) (n8n, 13 min)

## Frequently asked questions

**Is a knowledge graph the same as a database?**

It is a kind of database built around relationships. A spreadsheet stores rows; a graph stores things and how they connect.

**What is GraphRAG?**

An approach, published by Microsoft researchers, that builds a knowledge graph from documents and uses it to help a language model answer broad questions.

**Do I need a knowledge graph for a company chatbot?**

Usually not. A knowledge base of clear documents, with retrieval if it gets large, covers most office question-and-answer needs.

[All AI glossary terms, A to Z](https://learn.ynteractive.com/content/glossary) · [Free prompt templates](https://learn.ynteractive.com/content/prompts)