Pipeline to parse PDF files and convert them into structured Markdown using a combination of models that are time efficient.
This project is a high-performance pipeline that converts PDF documents into clean, structured Markdown using multiple specialized models. It supports extraction of:
Text blocks (via EasyOCR)
Tables (via pdfplumber)
Code blocks
Figures (via yifeihu/TF-ID large)
Hyperlinks & metadata (via PyMuPDF)
It’s designed for accuracy, modularity, and speed, making it ideal for large-scale document parsing or archiving workflows.