machine learning is a form of artificial intelligence where automation increases through experience and by the use of data in finance it typically means using statistical models to make predictions on possible outcomes or to improve forecasting in due app we use machine learning to improve automation by suggesting the probable workflow paths and the correct coding machine learning assists also with po matching and analyzes the figures in reporting here's an example i have machine learning predicting first the workflow based on the captured invoice metadata like vendor dates and keywords it checks the vendor invoice history for the possible patterns and then calculates the probability of different routes of approvers the invoice might travel you see here when i change the coding to machine learning suggested the prediction of the workflow changes this is logical as people usually represent different departments and cost centers so the coding prediction also changes accordingly you can also set a probability threshold so that after 90 probability the suggestion is applied automatically and ap doesn't have to set the suggestion coding predictions are not only one-liners but they might be complex and include multi-line predictions taking into account the invoice details and history after applying the machine learning suggestion you can still do changes and fine-tune it before sending it out do app learns the changes automatically and next time it'll be even better