Low accuracy of results

Help Document Understanding
1

Hello everyone,
I’m currently developing a robot that reads a pdf file with document understanding (with a form extractor) but on the template that I developed , the information that is presented/extracted in the validation station isn’t accurate, if the extractor works on the pdf I used as template it works with 100% accuracy but if I use another pdf with the same format/scheme it doesn’t work at all.

2

Hi, have you tried adding anchors to your form extractor?

3

Tried with anchors/custom area and even tokens I’m kinda lost about what can I try more.

4

Can you share your exported extractor and a sample document?

5

Says that I can’t share since I’m a new member

6

can I share it some other way?