Confluence Integration
Import pages and spaces from your Atlassian Confluence workspace to train your SiteGPT chatbot. Perfect for teams using Confluence for internal documentation, knowledge bases, and product wikis.Prerequisites
- An Atlassian Confluence Cloud account
- Read access to the spaces you want to import
- Owner or Editor permissions on the SiteGPT chatbot
Connecting Confluence
Authenticate
Click Connect Confluence and sign in with your Atlassian credentials. Grant SiteGPT permission to read your Confluence content.
Select Content
Browse your Confluence workspace and select:
- Entire spaces
- Specific pages
- Page trees (parent page and all children)
What Gets Imported
| Content Type | Included |
|---|---|
| Page text | ✅ |
| Headings & structure | ✅ |
| Tables | ✅ |
| Code blocks | ✅ |
| Attached PDFs | ✅ |
| Images | ❌ (alt text only) |
| Comments | ❌ |
| Page history | ❌ |
Only the current published version of each page is imported. Draft pages and page history are not included.
Best Practices
Choose the Right Spaces
Select spaces that contain customer-facing or support-relevant content:- Product documentation
- FAQ and troubleshooting guides
- Feature explanations
- Getting started guides
Exclude Internal Content
Avoid importing spaces with sensitive internal information:- HR policies (unless for internal chatbots)
- Financial documents
- Confidential project pages
Keep Content Fresh
- Re-import regularly: Sync your Confluence content periodically to capture updates
- Use labels: Tag pages you want to include for easier selection
- Clean up first: Archive or delete outdated pages before importing
Troubleshooting
Spaces not appearing
Spaces not appearing
- Verify you have read access to the space in Confluence
- Check that the space isn’t restricted to specific users
- Try disconnecting and reconnecting your Atlassian account
Import fails or times out
Import fails or times out
- Try importing smaller spaces or fewer pages at once
- Check for pages with very large attachments
- Ensure you have a stable internet connection
Content missing from imported pages
Content missing from imported pages
- Embedded macros may not render (content shows as placeholder)
- Images are not included (only alt text)
- Ensure the page is published, not in draft status
Confluence Data Center
If you need to train from Confluence Data Center, consider:- Exporting pages as PDF or Word and using file upload
- Using the API to export content to a supported format