Guides
Start typing to search…

Document Analyst - Overview

This guide explains what Document Analyst is, how it works, and the file types and limits for analyzing user-uploaded documents.

What is the Document Analyst feature?

Document Analyst is an advanced feature that allows users to upload files during conversations with your agent. The agent analyzes uploaded files against your knowledge base to provide contextual responses.

In simple terms: Users upload a document, ask questions about it, and your agent responds using both the uploaded content and your organization's knowledge base.

What are the supported file types?

Document Analyst accepts documents and images in the following formats:

  • Documents: PDF (.pdf), Word (.docx, .doc), OpenDocument (.odt), Plain text (.txt)
  • Images: JPG, JPEG, PNG, WebP

What are the file upload limits?

Premium Plans

File size: 5MB per file

Word count: 3,000 words total per prompt

Number of files: 1 file per prompt

Enterprise Plans

File size: 5MB per file (extended limits available)

Word count: 3,000 words total per prompt (extended limits available)

Number of files: 3 files per prompt (up to 10 available on request)

How multiple file upload works

Enterprise users can upload up to 3 files in a single message. When uploading multiple files:

  • All files must be valid format and under the size limit
  • Files are read in the order you upload them
  • The 3,000-word limit applies to all files combined
  • If any file is invalid, you cannot send the message

Important Notes

  • Word count limits apply to all uploaded files combined, not per file
  • Files are processed in upload order until the word limit is reached
  • Content beyond the limit is automatically truncated

🚧

Enterprise customers can request extended limits for file size, word count, or number of files through their account manager.

How It Works

  • Users click the attachment icon next to the chat input field and select their file(s). Enterprise users can upload multiple files at once
  • The agent processes files in the order uploaded. For multiple files, each file includes a header showing its filename so the agent understands it's working with multiple documents.
  • The agent performs multi-step reasoning, combining the uploaded document's content with your knowledge base. This allows the agent to provide richer, more contextual answers.
  • The agent delivers a response that integrates information from both the uploaded document and your knowledge base, with sources cited.

Advanced Vision Intelligence

When users upload images, Document Analyst understands visual content contextually:

  • Analyze architectural plans and blueprints
  • Compare product photos against specifications
  • Interpret technical diagrams and schematics
  • Review screenshots and error messages
  • Match images against your knowledge base

If your knowledge base includes images, Document Analyst can compare uploaded images to find matches or similar items.

Security

  • Files are processed securely and stored temporarily during the session only
  • Files are not added to your agent's knowledge base
  • Files are not shared with other users
  • Only the user who uploaded the file can access it during their conversation

Availability

Document Analyst is available on:

  • Premium plans (with standard limits)
  • Enterprise plans (with custom limits available)

Document Analyst works across all deployment types: web, mobile, embedded widgets, Slack, and API.

Related articles


Updated 4 days ago