PDF-to-RAG Readiness Scan

Parse a real PDF into page-level text, fragmentation, and extraction-order diagnostics before committing it to a RAG pipeline.

Which pages need OCR, layout recovery, or manual review before chunking?

Uses PDF.js to extract text items and coordinates page by page, then reports low-text, fragmentation, and reading-order inversion signals.

The selected PDF is sent to Fanout only for this no-store parsing request. It is parsed in memory and not persisted or sent to an AI model.