Import from website URL
Goal: Index one public web page without crawling the whole site.
When to use URL import vs crawl
Section titled “When to use URL import vs crawl”| Approach | Best for |
|---|---|
| Import URL (this guide) | A single FAQ page, blog post, or policy page |
| Website crawl | Discovering many pages across a domain — Website crawl |
Step 1 — Open import
Section titled “Step 1 — Open import”- Go to Knowledge bases and open your knowledge base.
- Click Import URL (or Add content → URL).
- Paste the full URL, including
https://.
Step 2 — Configure options
Section titled “Step 2 — Configure options”| Option | Guidance |
|---|---|
| Classification | Set public, internal, confidential, or restricted |
| Page title | Auto-detected; override if needed for your team |
Chatevo fetches the page, extracts readable text from HTML, and queues it for indexing.
Step 3 — Wait for indexing
Section titled “Step 3 — Wait for indexing”The imported page appears as a document with status Processing, then Indexed when ready. A snapshot captures the page content at import time — re-importing the same URL later creates a new snapshot if the live page changed.
Troubleshooting
Section titled “Troubleshooting”| Issue | Fix |
|---|---|
| Import fails | Confirm the URL is public (no login wall) |
| Empty content | Page may be JavaScript-only; export or upload a static HTML/PDF copy |
| Wrong text | Check for duplicate navigation chrome; prefer the main content URL |
Step 4 — Test retrieval
Section titled “Step 4 — Test retrieval”Attach the knowledge base to an assistant and ask a question specific to that page. Hybrid search (dense + keyword, merged with RRF) should surface relevant chunks. See Processing and indexing.
Related
Section titled “Related”- Website crawl
- Supported file types
- Auto-retrain — Standard+ plans can refresh URLs automatically