The Data Sources section in your Knowledge Base allows you to connect more than just articles. You can enhance your AI-powered experience by syncing external data like websites and documents to help AI Pilots and Copilot generate better answers.
- Users must have the Manage Knowledge Base permission.
- If the user does not have permission to manage the knowledge base and organize categories, the Knowledge Base section will not be visible in the Library Setup.
Topics Covered
Steps to Configure Data Sources
- Navigate to Setup
- From Setup, Go to Library >> Knowledge Base
- You will land on the Settings tab by default
- Select the Knowledge Base you wish to configure
- Click on Data Sources
Different Data Sources
You can configure the following data sources for your knowledge base.
Articles
This section provides quick access to your article listing page for the selected knowledge base.
Steps to Add and Configure Articles
- You can edit the article. Click here for steps on how to edit the article.
- Update the AI settings
- Enable its use in AI Agents and CoPilo
Website
You can sync content from your public website directly into your Knowledge Base to power your AI Pilot and Copilot. This feature helps you pull relevant content automatically from help centers, documentation, or any support-related pages using either a website URL or a sitemap.
Steps to Add and Configure Website URLs
- Click Enable Website Source.
- Click on Add to launch the setup popup.
-
Add Website via URL or Sitemap
- Website URL: Input the homepage or main support page of your site.
- Sitemap URL: Submit your sitemap URL for more structured importing.
- Usage Selection: Choose if the data should be available to AI Pilot, AI Copilot, or both.
-
Configure Advanced Settings
- Additional URLs: Add specific sub-pages that might not be discoverable via the main URL.
- URLs to Include: Add specific URLs to include them.
- URLs to Exclude: Add specific URLs to exclude them.
- CSS Selectors to Include: Only scrape sections matching the specified CSS selectors.
- CSS Selectors to Exclude: Avoid scraping UI elements like sidebars or headers.
- Clickable CSS Selectors: Used to simulate clicks and reveal hidden content before scraping.
- Wait-to-Load CSS Selector: Ensures scraping begins only after the specified element loads.
- Click on Save.
After saving, a table displays all added website URLs with their AI Pilot/Copilot status, domain, synced page count, and last synced date. Clicking the synced count opens a list showing each page’s sync status—Synced, Pending, or Failed—along with options to retry failed pages and view raw data for successful ones.
Note:
- Only public pages can be synced. Pages requiring login or human verification are not supported.
-
Sync limits:
- Up to 10 website sources
- Up to 3,000 pages per source
- Each page's text content must be under 100 KB (post-HTML strip).
Documents
Upload and manage PDF files to enrich your knowledge base content.
Steps to Add and Configure Documents
- Click Enable Documents Source.
- Click Upload PDF to add a new document from your system.
- Usage Selection: Choose if the data should be available to AI Pilot, AI Copilot, or both.
- Click on Save
Note: Documents are read-only and cannot be edited in the knowledge base.
This feature helps you scale your knowledge base by pulling in helpful content from multiple touchpoints, boosting your AI agents’ performance, and enriching the support experience.
Comments
0 comments
Article is closed for comments.