Chatleh logo markChatleh

Managing Your Chatbot's Knowledge Base

Understanding Website Scraping

When your chatbot is created or updated, it uses an intelligent scraping system to learn from your website:

  • Automatically follows internal links within your domain
  • Intelligently extracts meaningful content while ignoring UI elements
  • Processes up to 50 pages per scraping session
  • Handles JavaScript-rendered content

⚠️ Important: For best results, temporarily disable Cloudflare or similar bot protection during the initial scraping process. You can re-enable it afterward.

Knowledge Input Methods

Manual Data Input

Add knowledge directly through our text interface:

Google Workspace Integration

Import from Google Docs:

Import from Google Sheets:

Excluding Content

To prevent your chatbot from using certain content:

  1. Navigate to "Remove URLs from Knowledge Base"
  2. Enter the URLs you want to exclude (one per line)
  3. Click "Remove and Save Exclude URLs"

Common uses for exclusion:

  • User-generated content areas
  • Blog comments sections
  • Outdated documentation
  • Temporary promotional pages

Additional Knowledge Sources

Enhance your chatbot with knowledge beyond website content:

Custom Knowledge Base

Add information in the "Additional Knowledge" section using these formats:

Q: What are your business hours?
A: We are open Monday-Friday from 9 AM to 6 PM EST.

Q: Do you offer refunds?
A: Please contact our support team for refund inquiries.

Key Information:
- Customer support phone: (555) 123-4567
- Emergency contact: support@example.com
- Response time: Within 24 hours

Guided Prompts

Create specific responses for common queries:

  • Enter one prompt per line
  • Be specific and direct
  • Cover frequently asked questions
  • Include important business policies

Monitoring and Updates

Keep track of your chatbot's knowledge:

  • View all scraped URLs in the Knowledge Settings
  • Check last scraping date for each URL
  • Monitor training status in real-time
  • Retrain on updated content using "Scrape Current URL"

Best Practices

  • Regularly update your chatbot when website content changes
  • Use clear, concise language in additional knowledge
  • Test your chatbot after making significant changes
  • Keep excluded URLs list up to date