Cash Matching Automation using RPA & Machine Learning

1Rivet implemented a Azure PaaS based solution along with self-learning RPA using UiPath and Azure Machine learning to perform the activities related to the cash matching process.

This project utilized self-learning model, so if the system is not confident on identifying customer from wire instruction, it allows the Backoffice user to identify the customer. The system learns from this user intervention, so that the system can automatically detect the customer in the future.

Outcome

Set up Auto Cash App with the advancement of Technology which allows for both repetitive and complex procedures to be automated and reduce both the possibilities for human error and repetitive clearing process.
Leveraged Cloud Platform-based machine learning engine (Azure ML) to read the New incoming payment by different trustees and use the data set to predict matches for the correct customer and open receivable. This can be done either automatically cleared or suggested for review by accounts receivable.
Integrated RPA to match the cash with the open receivable in LoanIQ by Auto or user trigger option and update the underlying Systems/databases to get the real time results. PA Turns in Faster processing of incoming payments which reduces in the DSO (Days sales outstanding) and TCO (Total cost total cost of ownership).
Set up all-in one dashboard to monitor real-time Cash Collection and process of cash matched against open receivable.
Introduced Notification and Reporting facility for Executive and Operational levels to analyze Daily, Monthly, quarterly, and yearly Data which can leads to better cash forecasting and improved company planning/growth.
As a result, on Day 1, about 85% of the wires were automatically processed without any manual intervention. And as system learned from user intervention, the system was able to process 97% of the wires within 6 months of going live.