OCR technology without templates

I have one question about the OCR technology to be used without templates.
For example if I have 10 different vendors, they have 10 different ways of document layout. How we can extract the information from there and store it without to create templates for vendor 1 with regex, then for vendor 2 different regex and so on…?

Do you believe is it possible some kind of automation like this or at least creating a way harder logic in the robot to reach at least 60-70 success rate of extracting with only one template from all different invoices.

Thanks in advance!

That’s unfortunately the way I’ve been doing for a long time. Very time consuming and I promise you that you will soon discover that your vendors change the formatting from time to time.
I would advice using the new machine learning function in UiPath. I suppose it’s slightly more complicated to start with but in the long run you will most likely thank me :slight_smile:

1 Like