Your intelligent bot to chat with your documents

After a question has been entered in the amberSearch search line and the search results are displayed, you can click on the amberAI symbol for a desired search result.

This will take you to the amberAI Document Chat.

In the amberAI Document Chat window, the selected document is displayed on the left side and the amberAI Document Chat on the right side.

You can now enter specific questions about the open document in the amberAI Document Chat-bar and the AI will answer them.

Common Problems:

🚨 Error: Document not indexed

What it means
In document chat, the main document (and/or thread documents) don’t have indexed text. This is an indexing/content issue, not a general LLM failure. Since document chat is being phased out, you might rarely encounter this message. For the underlying concept of “chat with your documents” in amber, see the Help Center overview of the intelligent bot for documents .

Common causes

  • The file was uploaded but its text could not be extracted (e.g., scanned image-only PDF without OCR, protected or corrupted file).

  • The document’s indexing job failed or is still in progress.

  • The document was removed or replaced, and the current chat references a non-indexed version.

  • The document/chat feature path you’re using is deprecated and no longer indexes new content.

Immediate actions (Quick Fix)

  • Check if the document is text-extractable:

    • Open the file locally and try to select/copy text. If not, convert to a text-based PDF or DOCX (apply OCR for scans) and re-upload.

  • Use the Upload functionality in the Chat Interface

  • Use Ctrl+A for manually selecting the Text from the Document in an Editor and copy paste it into the Chat

  • If you pasted very large documents, split them into smaller, text-based parts.

If the message appears repeatedly

  • Verify that the document is accessible and still present in the connected source.

  • Convert problematic formats (scanned PDFs, image-heavy slides) to text-based formats with OCR, then re-upload.

  • Contact support (support@amberoad.ai) if:

    • Multiple documents from the same source fail

FAQ

  • Is this a model outage? No. It’s an indexing/content problem with the specific document(s).

  • Are my permissions involved? Not typically—this is about missing/failed text extraction or indexing.

  • Can I still get help from amber? Yes. Use amberSearch to find the document and ask amberAI questions referencing its content, or paste the key passages