Back to Knowledge Base

Chat Interactions

Upload Document to Chat Context

Overview

Flow ID: upload-document-to-chat
Category: Chat Interactions
Estimated Duration: 1-3 minutes
User Role: All Users
Complexity: Moderate

Purpose: Upload documents (PDF, DOCX, TXT, CSV) to chat input to include their content in your message. System extracts text from files and combines with your typed message, allowing AI to reference document content. Up to 10 documents can be added to a single message.

Related Flows

Prerequisites

Before starting, users must have:

  • Active chat open
  • Document files to upload
  • Files under size limit (~2000 characters extracted text)

Step-by-Step Flow

Main Path (Happy Path)

Step 1: Locate Upload Button

  • User Action: In chat input area, find paperclip or upload icon
  • System Response: N/A
  • UI Elements Visible:
    • Chat input text area
    • Upload icon button (paperclip)
    • Other controls (send, dataset toggle)

Step 2: Click Upload Icon

  • User Action: Click paperclip/upload icon
  • System Response: File browser opens
  • UI Elements Visible:
    • Operating system file selector
    • File type filter (PDF, DOCX, TXT, CSV)

Step 3: Select Document File

  • User Action: Navigate to file, select, click "Open"
  • System Response:
    • File uploads
    • Text extraction begins
    • Progress indicator appears
  • UI Elements Visible:
    • Upload progress indicator or tooltip
    • Status: "Uploading..." then "Extracting text..."
    • May show filename

Step 4: Text Extraction Completes

  • User Action: Wait briefly (usually seconds)
  • System Response:
    • Text extracted from file
    • Document "pill" or badge appears in input area
    • Character count updates to include document text
  • UI Elements Visible:
    • Document pill/badge showing:
      • Filename
      • X button to remove
    • Character counter increases
    • Document count (e.g., "1 document")

Step 5: Optional - Add More Documents

  • User Action: Click upload icon again to add additional files (up to 10 total)
  • System Response: Each document adds as separate pill
  • UI Elements Visible:
    • Multiple document pills
    • Counter: "3 documents"
    • Character total includes all documents

Step 6: Type Message About Documents

  • User Action: Type question or instruction related to uploaded documents
  • System Response: Text appears in input
  • UI Elements Visible:
    • Text input with message
    • Document pills remain visible
    • Total character count shown

Step 7: Send Message with Documents

  • User Action: Click send button
  • System Response:
    • Message sent with both typed text and document contents
    • User message shows typed text
    • AI receives full context including documents
    • Documents listed in message as "Reference Documents:"
  • UI Elements Visible:
    • User message in chat
    • Document references shown
    • AI generates response considering documents

Step 8: Optional - Remove Document

  • User Action: Before sending, click X on document pill to remove it
  • System Response:
    • Document removed
    • Pill disappears
    • Character count decreases

Final Step: Documents Included in Conversation

  • Success Indicator:
    • AI response references document content
    • Documents successfully processed
    • Can continue with document context
  • System State Change:
    • Document content included in message
    • AI has access to file information
    • Message history includes document reference
  • Next Possible Actions:
    • Send follow-up questions about documents
    • Upload different documents
    • Continue conversation with document context

Alternative Paths & Strategies

Strategy A: Drag and Drop Files

When to use: Convenient file access

Steps:

  1. Open file manager alongside browser
  2. Drag document file to chat input area
  3. Drop file
  4. Extraction proceeds automatically
  5. Continue as normal

QA Note: Drag-drop not confirmed but common UX pattern. May not be implemented.

Error States & Recovery

Error 1: File Too Large

Cause: Extracted text exceeds 2000 character limit
User Experience:

  • Error: "File too large" or "Exceeds character limit"
  • Document not added

Recovery Steps:

  1. Use smaller document
  2. Extract relevant sections manually
  3. Create dataset instead for large documents

Error 2: Cannot Extract Text

Cause: Image-based PDF, corrupted file, or unsupported format
User Experience:

  • Error: "Text extraction failed"
  • Document pill shows error state

Recovery Steps:

  1. Remove failed document
  2. Convert to text-based format
  3. Use OCR for scanned PDFs
  4. Try different file

Error 3: 10 Document Limit Reached

Cause: Already have 10 documents attached
User Experience:

  • Error: "Maximum 10 documents"
  • Cannot add more

Recovery Steps:

  1. Remove less important documents
  2. Combine information manually
  3. Send multiple messages
  4. Create dataset for large document sets

Version History

Date Version Author Changes
2025-10-04 1.1 Iternal Technologies Initial comprehensive documentation

Notes

Important Considerations:

  • Documents are temporary; included only in current message
  • Text extraction happens automatically
  • Supports PDF, DOCX, TXT, CSV
  • Maximum 10 documents per message
  • Each document limited to ~2000 characters extracted text
  • For large or permanent document sets, use datasets instead

Use Cases:

  • Ask AI to summarize a report
  • Compare multiple documents
  • Extract specific information from files
  • Analyze document content

Best Practices:

  • Use for one-off document references
  • For frequent use, create dataset instead
  • Verify extraction succeeded (document pill appears)
  • Keep documents concise for better results

Common User Questions:

  • "Can I use same document in multiple messages?" - Must re-upload each time
  • "What if file is too large?" - Use datasets feature for large documents
  • "Does AI remember documents from previous messages?" - No, include in each message if needed
  • "Can I see extracted text?" - Currently no preview; improvement opportunity

Trigger

What initiates this flow:

  • User manually initiates

Specific trigger: User wants AI to reference specific document content without creating permanent dataset.

User Intent Analysis

Primary Intent

Include document content in chat message for AI to reference, analyze, or discuss.

Secondary Intents

  • Get AI to summarize documents
  • Ask questions about document content
  • Compare multiple documents
  • Extract information from files
Free Trial

Download Blockify for your PC

Experience our 100% Local and Secure AI-powered chat application on your Windows PC

✓ 100% Local and Secure ✓ Windows 10/11 Support ✓ Requires GPU or Intel Ultra CPU
Start AirgapAI Free Trial
Free Trial

Try Blockify via API or Run it Yourself

Run a full powered version of Blockify via API or on your own AI Server, requires Intel Xeon or Intel/NVIDIA/AMD GPUs

✓ Cloud API or 100% Local ✓ Fine Tuned LLMs ✓ Immediate Value
Start Blockify API Free Trial