Skip to main content

Import Document Files (DOC, DOCX, PDF)

info

AI SmartTalk supports the import of unstructured documents such as Word files and PDFs, automatically converting them into smaller, searchable knowledge chunks for chatbot integration.

Supported File Types​

  • DOC/DOCX: Upload Word documents with text-based content.
  • PDF: Upload PDFs containing readable text (non-scanned images).
warning

Ensure documents are properly formatted, as poorly structured files may lead to errors during processing.

How Document Processing Works​

  1. Chunking: SmartTalk divides the document into smaller, logical sections based on headings, paragraphs, or sentences.
  2. Indexing: Each chunk is indexed for quick retrieval by the chatbot.
  3. Categorization: Tags or categories can be added to improve searchability.

Steps to Import Documents​

Step 1: Upload the File​

  1. Navigate to the Knowledge Base section in AI SmartTalk.
  2. Click on Upload File and select your document.
  3. Review the file details:
    • File name.
    • Size (maximum 25 MB).
    • File type.
  4. Click Upload to initiate processing.

Step 2: Configure Settings​

  1. Assign tags to categorize the content (e.g., "User Manuals," "Policies").
  2. Mark the document as Active to make it available for chatbot queries.

Step 3: Verify Content​

  1. Open the document preview in the Knowledge Base.
  2. Ensure all sections are processed correctly:
    • Headings and paragraphs are intact.
    • No missing or corrupted content.
tip

Use clear headings and consistent formatting in your documents for better chunking results.

Example Use Case​

Scenario: Uploading a User Manual​

  1. Prepare a DOCX file containing a product user manual with the following sections:
    • Introduction
    • Features
    • How to Use the Product
  2. Upload the file to the Knowledge Base.
  3. Assign tags such as "Product Manual" and "How-To Guides."
  4. Verify that the chatbot can retrieve specific sections (e.g., "How do I use the product?").

Troubleshooting​

Common Issues​

  • Error: Unsupported Format

    • Ensure the document is in .docx or .pdf format with readable text.

  • Error: File Size Exceeded

    • Compress the document or split it into smaller files.
danger

Avoid uploading scanned PDFs or documents with heavily stylized text, as these may not process correctly.

Next Steps​

After importing your documents: