MedPDF Structurer

Turn chart-heavy professional PDFs into queryable structured knowledge

Problem

A developer trying to structure the Chinese medical textbook Internal Medicine found that after splitting by table of contents, he got stuck on parsing images, tables, and flowcharts. AI-generated parsing scripts kept failing. In medical texts these diagrams carry diagnostic logic that is lost if dropped.

Solution

Upload a PDF, automatically split by chapter, extract tables/flowcharts/diagrams as structured fields. Provide disease-centric knowledge card templates and export JSON/CSV for downstream training or retrieval. Mark unparseable diagrams for human review rather than silently dropping them.

Users

Medical educators, developers building medical knowledge bases or question banks, med students preparing for exams

MVP

Only do chapter splitting + diagram OCR + table extraction; export disease cards; highlight unresolved areas; support local-only processing of sensitive material

Revenue

Per-document purchase or subscription; API billing for training institutions and med-exam platforms

Why now

LLM layout understanding is just good enough, but general tools still lose key diagrams in professional textbooks — a vertical gap an indie dev can own