SearchBook a call

OPEN SOURCE / Applied AI

terbium

Parse documents into rows, using a model if unsure.

A Portable Document Format (PDF) file keeps its words but loses its structure. terbium rebuilds the tables in a PDF from the position of every word, reads slide, spreadsheet and image files into the same shape, scores its confidence on each table, and sends only the tables it cannot resolve to a model.

Language
Python
License
MIT
Research area
Applied AI

This project has received support from NodeMaven ↗. Sponsorship is a disclosed commercial relationship.