---
title: "Keep Bot Data Separated"
description: "Use each chatbot's own configuration and generated identifiers."
category: "Security"
canonical: "https://www.proxyai.app/help/security/keep-bot-data-separated"
---

> For the complete documentation index, see [llms.txt](https://www.proxyai.app/llms.txt). Request public content pages with `Accept: text/markdown`.

# Keep Bot Data Separated

Each chatbot has its own ID, knowledge namespace, channel settings, embed snippet, usage log, and agent inbox. Pick the intended bot in the sidebar's chatbot menu and copy embed code from **Configurations → Channels → Web**, and upload documents through that bot's **Bot Knowledge** tab. Do not reuse identifiers or snippets from another chatbot.

When scanned-PDF enhancement is enabled, OCR runs on-device and documents are not sent to an OCR service.

### Check the bot before copying or uploading

Treat the selected chatbot as the boundary for a setup task. For example, an assistant answering questions for one store should receive that store's policies and the embed snippet generated for that bot. A second business or test assistant should not accidentally receive the first one's documents simply because both are visible in the same dashboard account.

Before an upload, check the bot name and the document's audience. After implantation, review the knowledge list in that same bot and test a question whose answer should come only from the new document. Before installing a snippet, compare it with the code shown in the selected bot's Web channel settings.

When reviewing conversations or costs, use the same bot selection so the evidence refers to the configuration you are investigating. The [Usage Log guide](/help/quick-faqs/how-do-i-review-or-export-usage) explains the available filters and export.

If information appears in the wrong assistant, record which bot, document and channel were involved before making further changes. Avoid copying additional material between bots while trying to establish what happened.
