> ## Documentation Index
> Fetch the complete documentation index at: https://docs.a1facts.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Knowledge Graph

> The knowledge graph is a dynamic, queryable database of your structured information, designed to be accessed through natural language.

The knowledge graph is where your structured data, defined by the ontology, is stored. `a1facts` uses a graph database to store entities and their relationships, allowing for complex queries and analysis. It is the central repository of factual information that your AI agents can rely on.

## Querying the Graph

The primary way to interact with the knowledge graph is through natural language queries. When an agent receives a query, it uses the `query_tool` (see [API Reference](/api-reference/knowledge-base)) to translate the natural language query into a precise graph traversal. The results are then returned to the agent.

<RequestExample>
  ```python theme={null}
  # Example of an agent querying the knowledge graph
  result = agent.run("What products does Apple offer?")
  ```
</RequestExample>

<ResponseExample>
  ```text theme={null}
  Apple offers the following products: iPhone, iPad, Mac.
  ```
</ResponseExample>

This process ensures that the agent's responses are grounded in the factual data stored in the knowledge graph, eliminating hallucinations and providing precise, verifiable answers.

## Neo4j Backend

For robust and scalable deployments, `a1facts` supports [Neo4j](https://neo4j.com/) as a backend for the knowledge graph.

<Tip>
  When you initialize a `KnowledgeBase` with `use_neo4j=True`, `a1facts` will connect to your Neo4j instance and store the data there. This is the recommended approach for production environments or large datasets. If `use_neo4j` is `False`, an in-memory graph will be used by default.
</Tip>