I am using Custom AI's invoice processing model to extract who the Recipient("John Appleseed") is.
The Recipient of the invoice is always going to be one of the names in our employees list. And I think the AI model, and other Flows in general, could greatly increase their accuracy and usefulness if it could leverage that kind of information.
Is there a way to do this?
Thanks,
Solved! Go to Solution.
This is exactly what we working on.
We are leveraging NLP to match metadata recorded in your system with variantq that comes from other systems like the ones of your vendors.
This would work only if the name in your document exactly matches the name in your employee list.
We are working on providing metadata fuzzy matching capabilities but it will take a couple of months before it is available.
Thanks CedrickB,
By the way, below are some of the example I am talking about:
Answer: "China Ventures, L.P."
Extracted "China Ventures L.P. Attention: Mr John" -> extracted too much
Extracted "China Ventures, Limited Partnership" -> two ways of referring same name.
This is exactly what we working on.
We are leveraging NLP to match metadata recorded in your system with variantq that comes from other systems like the ones of your vendors.
Thanks Cedric!