Knowledge Base

Add files

Upload documents and turn them into indexed source chunks.

Availableknowledge-base

File ingestion supports common support-doc formats such as PDF, DOCX, Markdown, TXT, CSV, and HTML. Text extraction metadata is stored so weak extractions can be diagnosed later.

Scanned PDFs

Ochre can detect likely scanned PDFs, but OCR processing is not currently configured. Upload text-based PDFs or source documents when possible.