The challenge

For corporates in South Korea, the process of accounts reconciliation is a daunting task and UPS were no different.

Not only are payments collected from numerous sources across multiple banking platforms, they are often received in partial sums and frequently miss key identifying information such as remitter names. This means the work of matching collections to invoices is an onerous one and it is common practice for firms to employ staff solely dedicated to manually reconciling accounts.

As a multinational firm operating in South Korea, UPS faced an additional challenge whereby it needed to maintain four domestic bank accounts in a market where around 90% of payments and collections from local corporates are in local currency. As such, the logistics giant had two full-time staff spending days manually reconciling more than 5,000 payments it received monthly via the four local banks, resulting in the inefficient use of resources and also increased costs.

The solution

UPS worked with its bank, J.P. Morgan and implemented the smARt+ solution, developed in collaboration with Kiwoong, a fintech partner. The solution is a unique, first-of-its-kind multibank reconciliation tool which was developed specially to reduce the pain points UPS faced around reconciliation. Leveraging robotics process automation (RPA), the solution works by gathering collection information from customers’ local banks as well as invoice details from the tax authorities. Customer payments are further matched to invoices through machine learning which recognises patterns over time to accurately reconcile receivables.

The powerful solution requires the simple installation of a software application on UPS’s system, which enables the firm to retain confidential information around their collections and invoices.

“With the deployment of smARt+, we can expect to reduce manual reconciliation by 90% and 19 man hours. With the help of machine learning capabilities, we also expect to continue enhancing the match rate to reach 95% three months after going live with the solution,” recalls Monika Nelwan, Asia Treasury Director. “In short, we can now make a variety of predictions based on patterns without human intervention.”

Best practice and innovation

The smARt+ solution is a unique, first-of-its-kind multibank reconciliation tool which was developed specifically to reduce the pain points around reconciliation. It is aimed at reducing the largely manual process of reconciling accounts receivables that come from numerous sources such as bank accounts, electronic promissory notes, B2B banking products, credit cards and direct debits. Combining robotics technology which matches payments received to invoices, and machine learning that recognises transaction patterns, the solution can accurately and automatically reconcile payments as well as provide timely reporting and cash flow forecasting capabilities.

Key benefits

• Staff cost savings – 19 hours each month.
• Human errors reduced.
• Increased automation.
• Improved cash flow forecasting.
• Enables better resource allocation to more value-add roles.
• Quick and seamless implementation.

“The solution implementation was quick and seamless, taking just two to three weeks, with minimal disruptions to our day-to-day treasury activities. The machine learning capability will continue to recognise patterns as data builds, which means improved accuracy with cash flow forecasting over time,” concludes Nelwan.