Add content to the corpus — everything is chunked, embedded and searchable immediately.
Add a document
Bulk upload
A .zip of .pdf / .txt / .md / .docx / .pptx files — each ingested separately. Runs in the background.
Crawl a website
Ingests every page under the path you enter, follows links deep, and pulls in linked PDF / Word / PowerPoint on the same domain — the whole section (safety ceiling default 1000, raisable). Respects robots.txt unless overridden.
Automatic refresh & maintenance
Web sources re-ingest on each source's interval, at the off-peak time, only when idle. "Re-index built-in sources" refreshes the bundled regulatory corpus.