KeyBank projects that 450 full-time equivalent (FTE) worker processes will be automated by January 2022, requiring the $176 billion bank to augment its automation team. 

The Ohio-based bank, which in 2017 began to roll out robotic process automation (RPA) in its frontline, middle and back-office operations, currently executes about 300 FTE worker processes daily. The bank’s bots are deployed in commercial credit, consumer lending, commercial lending, the contact center, ACH payments, human resources and finance.  

KeyBank is actively recruiting to add to the bank’s dedicated RPA and process mining teams, Dominic Cugini, chief information officer of service digitization at the bank told Bank Automation News. “We’re going to add a squad of 10 people and they’re going to be focused on automation and low code [platform] to start,” he noted. “Low code is the next thing and, between low code and automation, that will be a very powerful play.”

Low-code platforms are those that require little to no coding. The bank is looking to select a low-code enterprise platform in the next 60 days to handle internal workflow and client originations for certain products, Cugini said.

“What I’m looking to do with automation is continue to digitize tasks, but do it intelligently,” Cugini told BAN. However, he cautioned that banks need to take a “holistic approach” to automation to achieve their desired cost savings and ultimate business goals. 

This process at KeyBank is a multistep one. “First, we get ideas from across the enterprise, where people might be looking at automation to help with their day-to-day job,” Cugini noted. “We train our team of process miners to look a little bit left and a little bit right of that specific area.” This helps avoid a frequent pain point for banks: automating a “bad task or a task that doesn’t even need to be done,” he said. Process mining is the analysis of the end-to-end business processes required ahead of new automation.

Reengineering any process also includes looking at core systems to learn whether they are being leveraged appropriately. The next step is to bring in optical character recognition (OCR) and low-code solutions, for example, to see if automation efforts can be taken a step further than simply process reengineering the core systems. 

KeyBank signed deals with Automation Anywhere in 2017, and with UiPath in 2020, to provide automation tools for its core banking processes. The bank uses Automation Anywhere for its RPA and document-processing tools, and uses UiPath’s RPA technology; Cugini noted that the bank is using both providers so as not to have “all our eggs in one basket.”  

“Right now, we’re creating a decision tree to see if one platform is a better fit in certain areas of business, in solving certain problems over the other,” Cugini said. “We’re going through modeling right now, but we have both Automation Anywhere and UiPath products at our disposal.”  

One of those solutions is Automation Anywhere’s IQ Bot, an AI-powered document data extraction tool that KeyBank is currently using in its front, middle and back office. The bot has helped the bank “get over some hurdles,” Cugini said, but added that there are other business use cases where it has not been as successful. “We’re going to look to bring in an even more powerful OCR tool to help us with those business cases,” he noted, with OCR tools from intelligent process automation solutions providers Abbyy and Kofax as two under consideration. 

While the current technology ramp-up has buy-in from the bank’s top leadership, Cugini said, he emphasized that automation remains an introspective process. One perennial blunder is automating “low hanging fruit,” simple tasks that never translate to a return on the investment, given the tech costs and bot licensing, not to mention the bots’ upkeep and support. “That’s why reengineering is so critical,” he told BAN. While Cugini noted KeyBank had seen significant savings since the RPA rollout, he would not give further details on the ROI. 

“I think a lot of people are looking at automation as that single silver bullet that solves all problems but, again, it’s thinking strategically about what digitization is, how to achieve it and what role automation plays in it,” Cugini noted. “Automation is a tool, not the solution.”