Comment by pumainmotion

Curious to know since you mentioned that it was fast for thousands of PDFs... any rough timing information on some of your queries for that kind of dataset?