Knowledge Sources

Your chatbot's knowledge comes from the sources you add to your project. Knest supports both URLs and file uploads to build your knowledge base.

Adding sources

Go to the Sources tab in your project dashboard to manage your knowledge base.

URLs

Add web page URLs to index your documentation, help articles, or any public content. Knest crawls each URL, extracts the text content, and makes it searchable.

  • Paste the full URL including https://
  • One URL per source (add multiple sources for multiple pages)
  • Click Sync to crawl and index the content

File uploads

Upload documents directly to include content that isn't available as a web page. This is useful for internal docs, PDFs, or other files.

  • Select or drag-and-drop files to upload
  • Content is automatically extracted and indexed

Supported content

Knest works with public URLs:

  • Documentation sites
  • Help centers and FAQ pages
  • Blog posts and articles
  • Any publicly accessible web page

You can also upload files directly (PDFs, text documents, etc.).

Note: Content must be public and accessible without authentication. Password-protected pages and content behind login walls aren't supported.

Keeping sources up to date

When your documentation changes, you should re-sync the affected sources:

  1. Go to the Sources tab
  2. Find the source that needs updating
  3. Click the Sync button to re-crawl and re-index

The chatbot will immediately use the updated content.

Best practices

Start with your most important pages

Add your FAQ page, getting started guide, and most-visited help articles first. These are likely what users will ask about most.

Add specific URLs, not just homepages

Knest indexes one page per URL. For comprehensive coverage, add each important article or doc page individually rather than just the homepage.

Review conversations

Check the Analytics tab to see what users are asking. If the chatbot can't answer certain questions, you may need to add more sources covering those topics.

Source status

Each source shows its current status:

  • Synced — Content has been indexed and is available to the chatbot
  • Syncing... — Currently crawling and indexing the content
  • Error — Something went wrong (check if the URL is accessible)
  • Not synced — Source was added but not yet indexed

Next steps