jev-information-extraction

An interactive PDF question-answering demo: it extracts text from a PDF, uses Jev to rank which passages actually answer a given natural-language question, and lets you inspect each candidate match's confidence and location on the page.

unverified — sourced from awesomejev.com; description independently written by AionEdge from the repo’s own GitHub metadata and README, not GitHub-API relevance-checked