RPA: What is it and could it boost your productivity?

Artificial intelligence (AI) has become an essential technology in the community banking industry. Banks are putting it to good use in regtech tools that make compliance teams more efficient and in lending solutions that quickly analyze applicant data. While AI has had considerable hype around it, one similar automation technology many community bankers may not have heard of is robotic process automation, or RPA. Think of RPA like a digital assembly line that powers through manual bank tasks.

“RPA is a set of technologies that can do repeatable, redundant tasks in an efficient, cost-effective manner without human intervention,” says Charles Potts, ICBA senior vice president and chief innovation officer. “Robotic process automation excels at processing structured data for high-volume, rule-based, repetitive processes that are prone to human error.”

Putting RPA to use

One of the best use cases highlighted over the past year was the deployment of RPA to help process Paycheck Protection Program (PPP) loans, Potts says. Flooded with documentation and applications for PPP loans, community banks faced the challenge of pumping all that data into the Small Business Association’s (SBA) loan system. Many community banks employed RPA technology to mimic the human data entry process with great success instead of manually entering the data.

Queensborough National Bank & Trust Company in Louisville, Ga., achieved impressive results using RPA for its PPP loan processing.

“We used [RPA] exclusively for PPP loan origination in 2020,” says Cliff McGahee, director of core solutions at the $1.8 billion-asset community bank. “We leveraged tools we had in-house to interact with our customers and to get data in front of loan officers to make decisions. We were able to originate 1,780 loans for a total of $150 million.” In 2021, the community bank processed close to 1,000 additional loans, making it the fifth-highest originator of PPP loans for banks headquartered in the state of Georgia, he adds.

RPA can automate almost any bank process, says Erik Fisher, RPA solutions engineer for HelpSystems, an automation and security solutions provider in Eden Prairie, Minn. It can also handle end-of-day processing, account closures, customer information file (CIF) updates on core banking systems and integrating newer systems with older legacy systems that may lack modern connectors. The uses don’t stop there. RPA can assist with debit and credit card fraud processing, know your customer (KYC) processes, ACH stop payment, priority management, payroll information managing and moving data from one source into core systems, Fisher adds.

Combining RPA and AI

While traditional RPA works best for structured data, AI excels at tackling unstructured or semi-structured data, simulating human intelligence processes using computer systems or machines. Many community banks use AI to provide personalized customer insights and recommendations, detect and prevent payments fraud, analyze loan applicant data, and provide 24/7 customer service using chatbots.

Often, the best way to use RPA and AI is to blend them, taking the best of both technologies.

“For example, within mortgage processing, RPA can automate interactions between applications, data entry and extraction, and email and [file transfer protocol] automation,” Fisher says. “Then, by leveraging AI and unassisted machine learning, lenders can apply intelligence to mortgage document processing, eliminating costly steps like document classification, manual data entry and document analysis.”

“The promise of RPA and AI, and the best way to use them, is to help automate, speed up and scale those core banking processes—lending, account opening, collections—with smaller teams, while also preserving quality and a great customer experience,” says Joseariel Gomez, CEO of Berkeley, Calif.-based Shastic, an ICBA ThinkTECH Accelerator alum whose RPA assistant for banks, Elle, employs RPA and AI technology. “This leads to lower operation costs and more revenue with higher gross margins and better service.”

According to Gomez, Shastic is seeing substantial savings on staffing expenses based on a combination of eliminating 15% of phone calls on loan applications in process and converting 15% of other calls into automated interactions through channels like text messaging. This reduces bank staff’s follow-up work by 30%, which represents significant savings.

Shastic has also applied RPA to accelerate communications with prospects and customers during collections, lending or account opening processes. “It takes most banks an average of two to three days to get hold of and collect specific information or documents from customers during the process,” Gomez says. “We are seeing that Elle is cutting that down to an average of 12 minutes using a combination of RPA and AI.”

In an economy where community banks are looking for new sources of non-interest income, ways to cut expenses and gain market share, RPA could save them time on the manual tasks that don’t require more complex machine learning technology. And that could further enable the high-tech, high-touch relationship banking community bank customers are looking for.