Training Your Chatbot
Train Your Chatbot on Website Content
Website Source
1
Go to Training & Content
In your chatbot dashboard, click Training & Content in the sidebar.
Click the + Add Sources button.
From the source type dropdown, select Website.
Enter your website URL in the Website URL field.
Click the Crawl Website button. AI Chatbot will scan your website and discover available pages. This may take a minute or two depending on the size of your website.
Selecting Pages to Train
1
Review discovered pages
Browse the list of pages found on your website.
Use the checkboxes to select which pages you want to train your chatbot on. You can select all or choose specific pages.
Click the Add Selected button to add the chosen pages as training sources.
Auto-Sync Website
Before crawling, check the Auto-sync website pages daily checkbox below the URL input field.
2
Set exclusion patterns (optional)
Open Advanced Options to add URL exclusion patterns. Any URLs matching these patterns will be skipped during crawling. For example, you can exclude /admin/* or /login.
3
Create auto-syncing website
Click the Create Auto-Syncing Website button. AI Chatbot will crawl your website, discover all pages, and set up daily automatic syncing.
- New pages are automatically discovered and added as training sources
- Removed pages are cleaned up and removed from your training data
- Exclusion patterns are applied on every sync cycle
Link Sources
In your chatbot dashboard, click Training & Content in the sidebar.
Click the + Add Sources button.
From the source type dropdown, select Links.
Enter the page URLs you want to train on in the Link URLs field. Separate each link by a new line.
Click the Fetch Links button to retrieve the content from each URL.
Selecting Links to Train
Check the list of URLs and their content sizes.
Use the checkboxes to select which links you want to train on.
Click the Add Selected button to add the chosen links as training sources.
Managing Your Trained Sources
After adding sources, they’ll appear in your Sources list with their training status.
Each source shows:
- Name: The page title
- URL: The source URL
- Type: Link icon for web sources
- Size: Content size (e.g., 5.8 KB)
- Status: Training status (Trained, Training, Pending, Error)
- Auto: Whether auto-sync is enabled
- Last Trained: When it was last processed
Keeping Content Updated
Use the checkboxes to select the sources you want to update.
Click the Retrain button to re-fetch and retrain on the latest content.
Best Practices
- Start with key pages: Train on your most important content first (FAQs, product pages, documentation)
- Review content quality: Ensure pages have meaningful text content, not just images or videos
- Remove duplicates: Avoid training on multiple pages with the same content
- Update regularly: Retrain sources when your website content changes significantly