I built Papero around a simple problem: extracting a PDF is not the same as understanding its structure. A lot of PDF pipelines work well until you encounter multi-column layouts, tables, formulas, figures, or documents where you need to know exactly where a piece of text came from. Papero keeps that structure instead of flattening everything into plain text. What it currently focuses on: • PDF → Markdown / JSON / Excel / Word • Reading-order preservation • Tables, formulas and figures • Bounding boxes and source locations • Structure-aware chunking for RAG • Batch-friendly document processing The RAG part is especially important to me: a retrieved chunk should not lose the information about where it came from in the original document. I'm sharing Papero to get feedback from people working with document AI, RAG and PDF pipelines. What would you want Papero to support next?

Papero

Lightweight PDF extraction for structured data & RAG

Lightweight PDF extraction for structured data & RAG