# Knowledge Base (AI)

> The set of files an AI looks through. Then it gives its answer.

**Field:** Artificial Intelligence  
**Category:** Retrieval & Search  
**Also known as:** Knowledge Base, RAG Knowledge Base

## Think of it like

It is like a shelf of books. The helper reads them, then answers you.

## Definition

A knowledge base is the set of documents an AI system is allowed to search. It can hold help pages, policies, product sheets or past tickets. In a RAG system, it is cut into chunks and indexed before anyone asks. At question time, the system pulls the best chunks from it. The answer can only be as good and as fresh as what is in it.

## Example

An airline's help bot uses a knowledge base of 1,200 help pages. When the baggage rules change, the team updates one page, and the bot's answers change the same day.

## Why it matters

It lets a model answer from your own, current facts. No retraining is needed to update it.

## Common mistake

Loading every file you own. Old and clashing pages lead to wrong answers, so keep it clean.

## Three principles

- **Your documents.** What the system may search.
- **Prepared ahead.** Chunked and indexed before questions.
- **Keep it fresh.** Old pages give old answers.

## Related terms

- [Retrieval-Augmented Generation (RAG)](https://learn.benibix.com/term/ai/retrieval-augmented-generation)
- [Chunking (RAG)](https://learn.benibix.com/term/ai/chunking)
- [Document Parsing](https://learn.benibix.com/term/ai/document-parsing)
- [Vector Database](https://learn.benibix.com/term/ai/vector-database)
- [Retriever (RAG)](https://learn.benibix.com/term/ai/retriever)

## Told apart from

- [Knowledge Base vs Knowledge Graph](https://learn.benibix.com/compare/knowledge-base-vs-knowledge-graph): People think a knowledge base and a knowledge graph are the same store, and they are not. A knowledge base is a set of documents that an AI system searches as text. A knowledge graph holds facts as things and named links between them. You read the first. You follow the lines of the second.

![Left, the knowledge base is prepared ahead: files are parsed to clean text, cut into chunks and indexed. Right, a question is rewritten, then the retriever searches the index and passes the top k chunks to the model, which writes the answer.](https://learn.benibix.com/diagrams/ai3-rag-pipeline.svg)

---

Source: https://learn.benibix.com/term/ai/knowledge-base
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-18
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
