Why PDF Extraction Breaks RAG — And What I Built to Fix It

Description and use

Papero preserves PDF structure, exports to Markdown, JSON, Excel, and Word, and generates metadata for RAG, improving retrieval accuracy.

Problem

Converting PDFs to plain text often loses tables, reading order, formulas, figures, and source locations, which harms Retrieval-Augmented Generation (RAG) pipelines.

Outcome

Papero maintains structural integrity of PDFs, enabling more accurate RAG chunking and retrieval.

Workflow

• 1. Input PDF into Papero
• 2. Papero parses document structure
• 3. Export content to desired format (Markdown, JSON, Excel, Word)
• 4. Generate RAG chunks with metadata (section, page, bounding-box)
• 5. Feed chunks into RAG pipeline

Evidence

When PDFs are converted to plain text, tables, reading order, formulas, figures and source locations can easily get lost — and that can hurt RAG pipelines.

Source

I built Papero, an open-source PDF extraction tool that preserves document structure and exports to Markdown, JSON, Excel and Word.

Source

It also provides structure-aware RAG chunks with section, page and bounding-box metadata.

Source