AI Machine Learning for High Value Payments Repair

Challenge

Real-time gross settlement (RTGS) and SWIFT payments often need manual intervention to repair incorrect or missing data, adding high cost for the bank and delays for the customer

Considerations

1. How to mitigate risk in RTGS? RTGS payments are often high-value payments for high-value customers, so issues need to be rectified, but manual intervention adds risk of human error

2. How to improve CX? Manual intervention adds delays and uncertainty into the process for time-critical payments

3. How to reduce costs? Manual intervention necessitates multiple rounds of edits and checks, increasing associated service costs exponentially 

Solution

1. Making better use of existing technologies, such as fuzzy logic, combined with ever-improving machine learning and artificial intelligence could enable the bank to leverage rules written across the entire payments chain.

The implementation of machine learning across consortium data could be used to build rules built on macro and micro trends analysis by segment, payment type, beneficiary or initiator that allow for auto-completion or augmentation of payments data in order to remove the need for manual intervention.

2. Leveraging new payment messaging types and standards for cross-border and real-time payments (e.g., SWIFT gpi or ISO 20022) allows for enriched data to be included at the payments initiation level.

The CX of inputting that data correctly could be improved with strong front-end innovation in the user interface, as well as leveraging the in-flight tracking available from these technologies to provide real-time updates on status, including push notifications to the customer, removing the “where is my payment” uncertainty from the equation.

3. Reducing the risk of human error and the need for support staff to respond to ad-hoc status requests from customers will reduce the cost to operate, allowing banks to redeploy human resources on tasks that add value to the customer and the bottom line.

As real-time payment rails increase their transaction limits, corporate banks need to increase their SLAs, improve the customer experience for RTGS and innovate on overlay services in order to protect these revenues.

CX optimization

Create an international know-beneficiary database for banks within the consortium to match against, a “closed-loop confirmation of payee,” especially for cross-border payments, to reduce errors and improve straight-through processing rates, reduce cost to operate and augment the CX within the front-end initiation portal.