Do not overthink it. Set up paperless, create a watched folder. Paperless does the OCR for you. Scan your stuff and check if it was scanned as you want it to be. If yes, drop it in the folder. Tag as you go, paperless will learn and tags will get more accurate. If something reaches a level where you can trust paperless to always tag it correctly, let it tag that type of thing completely automated.

And file away your scanned papers separately, because scanning old things takes a lot of time and will most likely not be done in a day or two. Even with a scanner which can pull through stacks of pages, you still have to check if every page really was scanned (scanners can pull in two pages at the same time, only one page will be scanned then) and you have to merge multi-page docs (or scan them that way immediately).