Knowledge Base Upload
In addition to automatically crawling your website, paid plans can upload documents directly from the dashboard to enrich the AI's knowledge base. Uploaded content is merged into the same RAG retrieval pool as crawled pages, so the AI answers from both sources seamlessly.
Who can use it
Document upload is available on Pro and Enterprise plans. Free plans see an upgrade prompt.
| Plan | Max files / domain | Max single file | Total storage / domain |
|---|---|---|---|
| Free | — | — | — |
| Pro | 20 | 10 MB | 50 MB |
| Enterprise | 100 | 20 MB | 500 MB |
Supported file types
txt, md, markdown, csv, json, log, pdf, docx (Word), xlsx (Excel), pptx (PowerPoint), html.
Scanned images / image-only PDFs are not OCR-processed — only embedded text is extracted.
How it works
- Go to Dashboard → Knowledge Base, pick the target domain, and upload a file.
- The backend extracts plain text from the file and stores it alongside crawled content for that domain.
- The knowledge base is re-indexed (chunked + embedded) automatically. The original binary file is not stored — only the extracted text is kept.
- Visitors immediately get answers drawing on the uploaded documents.
Re-crawling is safe
Uploaded documents are stored independently from crawled content. Re-crawling a domain (manual or scheduled) refreshes only the crawled pages — your uploaded documents are never lost.
Removing a document
Delete a document from the list and it is removed from the knowledge base and re-indexed right away.
Privacy
The backend is stateless and does not retain the original files. Only the extracted text is persisted per domain to power retrieval, isolated by domain like all other site content.