Hi all,

the posting validation feature in journals has served us well in the past, but it does not include any type of "is this a typical transaction" type of checks.

I propose a machine learning based validation feature, that checks the plausibility of a transaction to be posted, compared to past transactions and based on multiple criteria such as who is posting, time of the month and year, how high the amounts, accounts and dimensions used, etc.

This feature should ideally apply for all documents, not just GL journals; we should be able to check sales orders and vendor invoices as well, before posting.

Keywords: artificial intelligence, AI, Chat-GPT, GPT, machine learning, neural network