Question about OCR text

I just watched your vid about OCR, very cool. Question I have about training the app, is it specific to training for the same document design? For example fuel receipts are usually the same size and the print can be similar, but a hardware receipt is larger with more text. And, if you have a small business that for the most part will use the same vendors, thus the same receipt designs. Can you create a category that will show cells/fields specific to that design? Hope that makes sense. User chooses a Fuel and the fuel design fields show, vs Hardware, etc.? or will one set up columns work for all?


Adding @tony

Basically, subsets of training data… Yes, we have a table for receipts, but really we need to tell which receipts are all the same… And then the training goes against the subsets…

Got it, thanks Grant!

I don’t want you to be confused. I was simply agreeing that it would be cool to have, not that we do have it…

@Rod Right now the OCR tool only works for documents that have the same design. You’ll need to create separate models for separate document layouts.

1 Like

Sub types would be cool though huh! And a serious use case.

1 Like

Yes, If that feature comes I would implement this into our manufacturing facility to make sure labels are correct. We have many customers but they all have different label formats for shipping. would be awesome.

Subs would be cool. But, for a firm that uses the same vendor most of the time this should be helpful, especially since admin can run an import into Quickbooks