@simon considering that PDFs are still hard to work with (eg select/copy/paste) after many years, I do wonder if this is an area where AI can help (based on what a user can see on scree, what do you think they want to copy) rather than a traditional approach