I doubt there is any ‘solid framework’ for this kind of stuff. PDF format isn’t meant to be machine-friendly by design, and AFAIK there is no guaranteed way to parse arbitrary PDFs.
Personally, I’ve used yob/pdfreader package (my fork with small fixes: divan/pdfreader) for one task involving parsing specific set of PDFs. This package is quite old, non-idiomatic Go, but it has some nice design solutions, which I was able to use to make PDF-to-text converter on top of that. For me, it was enough, and, with some of heuristics, I was able to finally extract data from PDFs. Try it, may be it’ll work in your case as well.
Regarding making PDFs, I’ve used gofpdf package (mentioned earlier) a lot in production, and it’s quite good.
i don’t think that this is the way. pdf are not design to be an editable file and definitely not in this mode. i think that the best practice is to generate the final pdf document from your data set and/or forms.