AP Document Recognition didn't recognize the alternate ID after 25 documents

  • 17 February 2022
  • 2 replies

Userlevel 4


I recognized a Vendor invoice in the Incoming Documents and the system didn't map the alternate ID at the first time, so I have to do it manually (select the field alternate id and select the correspoding number in the pdf), then, I saved the document. I did this process +25 times using the same format and the system didn't learn the mapping and every time I have to do the mapping manually.

I also Rebuild the text index but it didn’t work.

How many times the system has to recognize the same format of invoices to learn a manually mapping?


p.s. I have a 2021 R2 version in SaaS.



Best answer by rodrigocid34 8 March 2022, 17:52

View original

2 replies

Userlevel 4


Just to letting know to all Community, Acumatica’s Product Team mentioned regarding this topic:

“Based on the information from the product team, the ML/AI is only set to work with the header portion of the document.”

We (as an Acumatica partner) submitted an idea for finishing this module, in here 


Please, submit your vote, it’s very important for everybody.



Hi @rodrigocid34 This was one of the features we looked forward to when implementing document recognition. We quickly found out that the system does not in fact “learn” and update alternate ID’s automatically. We have about 70,000 sku’s and updating alternate id’s manually is time consuiming and ineficcient. Please let me know if you’ve made any strides on the Incoming Documents engine.


About Acumatica ERP system
Acumatica Cloud ERP provides the best business management solution for transforming your company to thrive in the new digital economy. Built on a future-proof platform with open architecture for rapid integrations, scalability, and ease of use, Acumatica delivers unparalleled value to small and midmarket organizations. Connected Business. Delivered.
© 2008 — 2024  Acumatica, Inc. All rights reserved