> For the complete documentation index, see [llms.txt](https://docs.aisuru.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.aisuru.com/en/agent-training/document-import/how-document-import-works.md).

# How document import works

Importing documents into AIsuru is a powerful process that lets you quickly enrich your Agent's knowledge. It transforms your documents into a series of question and answer pairs, creating an interactive and queryable knowledge base.

Here is how it works.

### The import process

1. **Upload the document**: upload your document in the Import / Export section of your Agent
2. **Analysis and splitting**: AIsuru analyzes the document's content and breaks it into distinct paragraphs
3. **Generating question and answer pairs** ([contents](/en/agent-training/contents.md)): for each paragraph, AIsuru uses generative AI to generate one or more question and answer pairs. The number of contents varies based on the document and the [granularity setting](/en/agent-training/document-import/importing-text-documents.md). Contents are created in the Agent's language, the one that was set when the Agent was created

### Benefits of document import

* **Time savings**: quickly turn large amounts of information into knowledge your Agent can use
* **Consistency**: generated responses are based directly on your documents, ensuring accuracy and coherence
* **Flexibility**: you can import different types of documents to cover a wide range of topics

### Limitations to keep in mind

While document import is a powerful tool, there are some limitations worth knowing:

* **Images and videos**: currently, AIsuru can't interpret or process images and videos in documents; import focuses exclusively on text
* **Supported file types**: AIsuru primarily supports importing plain or semi-plain text files. Other formats (such as JSON) may require a different approach, like using [functions](/en/advanced-features/integrations/functions.md) or [MCP](/en/advanced-features/integrations/mcp/mcp-what-they-are-and-how-to-use-them.md)
* **Non selectable PDF content**: if you upload a PDF that contains mainly images or non-selectable text (like scanned documents), AIsuru won't be able to effectively extract and process the content

### Tips for effective importing

* Use **well structured documents with selectable text** for the best results
* After importing, **manually review and refine the generated contents** to ensure maximum quality and relevance

Document import is a great starting point for building your Agent's knowledge base. When you combine them with other training methods (like direct conversation and manual content creation) you can build a truly comprehensive, well informed AI assistant.
