Financial institutions including JPMorgan Chase, Truist Bank and Wells Fargo are using AI models to expedite administrative processes and organize unstructured data through UiPath.

Unstructured data from emails, texts and voice notes can provide value to a bank but might take time to collect and store, Gartner analyst Moutusi Sau told Bank Automation News. 

Financial institutions are increasingly looking for AI solutions from third parties to manage unstructured data, Bill Hincher, Americas financial services industry leader at UiPath, told BAN.  

“A major issue with any type of data or AI strategy is that currently data is very fragmented, [and] unstructured,” Jennifer Warren, head of global markets digital strategy at Barclays Investment Bank, said at Fintech Connect in June. “To take full advantage of these emerging technologies [like AI] and get the most out of it [banks need to have] structured data.” 

Banks are looking to automate unstructured data analysis and back-end processes which are most often done manually to simplify customer journeys, and enhance customer experience, Hincher told BAN. 

For example, loan origination has historically been difficult for banks to automate until recently, he said, noting that in some cases a loan application is 100+ pages and bank employees have had to pull the data out and enter it into origination systems.  

UiPath is using AI to extract unstructured data from a variety of sources, including freeform text, emails and raw data needed for processes, Hincher said. In fact, more complex AI and decision trees can potentially be used for loan decisioning — if the inputted data meets the predetermined criteria.

Now, a process that may have taken five hours manually can be completed in five minutes through UiPath, he added. This can help banks drive down mortgage origination timelines significantly from the current industry average of 45 to 60 days. 

“That’s really where we’ve been able to see a lot of uplift done using AI,” Hincher said. “The specialized AI [is able] to extract the data out of those documents, and then get it downstream into the underwriting systems and their decisioning systems.” 

UiPath clients include IBM, Capgemini and Ernst & Young.