Search across every ingested document — Word files, PDFs, HTML, notes — in one place.
Agents get the same corpus over MCP.