Skip to main content
A powerful set of scripts for standardizing large collections of books, papers + other published documents. We use a combination of local + cloud OCR, Vision Language Models (VLM) and Large Language Models (LLM) to extract and intelligently generate metadata + filenames. This project is a work-in-progress, as both a tool and a learning project for AI-led agentic development.

​
Before

├── Androids Dream of Electric Sheep__English-242L.pdf
├── Quantum Computing Introduction MITPRESS_2011.pdf
├── Complexity ihn Physics- .pdf
├── GOODFELLOW_AVIAN (books about birds).pdf
├── j.physrep.2024.01.012.pdf
└── 10.1007-978-3-031-04083-2.pdf

​
After

├── Do Androids Dream of Electric Sheep, (Philip K. Dick), Doubleday, (1968).pdf
├── A Gentle Introduction to Quantum Computing, (Eleanor Rieffel), MIT Press, (2011).pdf
├── More Than the Sum of the Parts, Complexity in Physics and Beyond, (Helmut Satz), Oxford University Press, (2022).pdf
├── Avian Architecture, (Peter Goodfellow), Princeton University Press, 2nd Ed, (2024).pdf
├── Quantum Phase Transitions in Driven Systems, (Smith et al.), Physical Review, (2024).pdf
└── Emergence in Complex Networks, (Lee Johnson), arXiv, (2024).pdf
This project is a work-in-progress:
  • Back up your PDFs
  • Run the scripts iteratively on a small subset of your collection before scaling up
  • Monitor your cloud API costs: Displayed costs are only estimates

Quickstart

Get up and running quickly

Key Concepts

The main ideas behind the project

Use Cases

Is this project right for you?