guys, I made a website, DullyPDF, where you can automatically convert a pdf to a fillable form. You take raw pdf and then a ML algorithm (jbarrow open source field detector) finds all the input areas like name, date, address etc and creates fillable form fields there. Then you can database map the fields allowing you to fill in any person/id information in automatically. Adobe has something similar but without the ability to rename based on a db and it cost 20$ a month. My site has everything Adobe has for free (with some limitations for free users because backend GPU cost would be too high, maximum 10 page detections), however you could host the UI + detection yourself and not worry about that.

So raw pdf -> db template + filling

It also has a great form builder UI, the detector outputs confidence values for how likely it is a field, and the DullyPDF UI categorizes with yellow for medium and red for low confidence. You have all PDF form builder functionality in this UI too. My UI built lightweight representations of field to deal with lag, most viewers use actual acro fields which causes too much lag past PDFs two pages or more. Once you want to download the PDF, then the actual fields get created.

Feel free to try it out for free, let me know if the software is helpful or not all feedback would be appreciated. I think Dental + Patient Intake forms work well is this site.

DullyPDF

Github