How to extract data from an unstructured pdf where lable are also missing

Hello everyone,
I have to process the invoices that came to me as a scanned copy in the pdf form. I need to extract around 15 fields from it. But the problem is that for many fields there is no lable mentioned in the pdf and also the pdfs are completely unstructured in some cases the information is at the top and in some cases it is in the middle. I am using google OCR for it. Please help me to sort out this challenge.

Thanks in advance!!

Without certain pattern or labels (relative scraping) it is difficult to extract the data.
Try to find out the different patterns for types of invoices and extract the data.

Basically, the scenarios which you explained is not correct one for automation RPA. as the possible outcome will not be good.