🚀 go-pugleaf

Consider the following variables before accepting any claim of a given throughput rate:
scanned image dpi - 200 vs 300 vs 400
image processing - pre or part of ocr processing
b/w vs color
text based images vs graphics heavy files
zoning
text flow (column) recognition
machine resources - cpu and RAM
network processing path/traffic
output format (Searchable Image, Formated Text and graphics, etc.)
compliance with PDF specification
PDF version
optimization/compression of images, and of PDF files

I do not claim that the above is an exhaustive list of every consideration.

Your usage requirements may affect choices among those (and other) parameters, in turn affecting your throughput. For example, a smaller file size comes with Formatted Text and Graphics, but Searchable Image is more likely to comply with minimum specified standards for legal or archival usage.

BTW, is the quoted 8,000 per day based on hourly for 24 hours per day, or an 8-hour shift? Ask them what the specs of the input files, output files, processing station are to achieve that rate.

Noah Katz