# Knowledge Pre-processing

Raw data isn't always "AI-ready." Your knowledge base might contain thousands of pages originally written for human readers, PDFs, or internal databases. Pre-processing is the bridge that translates that static information into a format that makes your AI agent faster, more accurate, and more helpful.

## Why you should pre-process knowledge

* **Ensuring "Chat-First" Clarity**: Traditional documentation often tells users to "call us" or "click the button below." Since your AI lives in a chat interface, pre-processing allows you to swap those instructions for chat-specific actions, like "I can help you right here."
* **Improving Accuracy (Signal vs. Noise)**: By using tools like Truncate, you remove "filler" text and legal jargon that might distract the AI. This ensures the agent focuses only on the most relevant information, leading to fewer hallucinations and more direct answers.
* **Future-Proofing Your Data**: Instead of manually editing 500 uploaded documents when a product name or URL changes, you can use a Search & Replace filter. This acts as a global "Find and Replace" that keeps your AI’s knowledge current without requiring a single file re-upload.
* **Consistent Brand Persona**: Using Append or Prepend allows you to "wrap" every piece of data in your brand’s voice. You can automatically add context or disclaimers to every response, ensuring the AI always stays within its operational guardrails.

## When you should pre-process knowledge

If you find your AI agent is consistently giving the "right" answer but with the "wrong" instructions (like giving an old phone number), instead of changing your original documents, simply add a Source Filter. It’s the fastest way to fix AI behavior at scale, and sometime the only solution if the original source can't be changed.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.ebbot.ai/ebbot-docs/core-capabilities/ebbotgpt/ebbotgpt-knowledge/knowledge-pre-processing.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.