Documents & Data

What Ragbee ingests, how it's indexed, and how to control retention.

Supported sources

• Direct upload — Dashboard → Documents. PDFs, Word and text documents, spreadsheets and other common office formats
• URL ingest — paste a page URL and Ragbee extracts and indexes its content
• Connectors — Dashboard → Connectors links cloud accounts to import from: Google Drive, Dropbox, Microsoft OneDrive, Atlassian (Confluence)

How documents become answers

1. Each document is split into chunks sized for retrieval
2. Chunks are indexed per workspace — workspaces keep document sets separate (e.g. "Employee Handbook" vs "Product Docs"), and each bot or chat session answers only from the workspace it's pointed at
3. When someone asks a question, Ragbee retrieves the most relevant chunks and writes a grounded answer with the source passage cited — if nothing matches, it says so instead of guessing

You can inspect exactly what was extracted: the document list shows chunk counts, and document detail views show the chunk contents — no black box.

Managing documents

• Delete removes a document and its chunks from the workspace
• Restore is available for recently deleted documents; purge removes them permanently
• Re-upload to update: the new version replaces the old one's chunks in that workspace

Retention & data control

• Chat and chatbot conversations, events, and transcripts are visible in your dashboard for audit, and retention purging can be scheduled per tenant (ask support to configure)
• Personal data in public chat channels is minimized by policy — see Security & Policies
• Tenant deletion (all documents, logs, and personal data) can be requested at support@ragbee.ai