Anil Khatri has been with SAP for the last 18 years. In his current role he is head of Information Technology at SAP for South Asia.
SAP developed a Cash Application which is a new machine learning solution based on SAP Leonardo. The application has been invented to improve and optimize the level of automation for matching incoming bank receipts to open customer invoices for the companies. The solution is highly integrated into the existing SAP environment and simplifies the daily work of Global Finance Shared Services (GFSS) at SAP. Historical data from the ERP is used to train the Cash Application machine learning model in the Cloud.
The trained model will later be used to make matching predictions on the bank statements and open invoices. Bank statements and open invoices from the ERP are sent to Cash Application on Cloud for a predictive matching. The matching proposal from Cash Application will then be returned to the ERP. This match can also be set for automatic posting, according to a confidence threshold defined. Otherwise, a report will be displayed with the matching proposal.
Through the application, automating a part of the accounts receivable process was achieved. It also lead to rule based programming that can only match the invoice to the bank statement when the required information is present. When information is missing, the automation fails, and an accountant will have to manually find the Invoice. Cash Application is designed to find the matching invoice despite the missing information.