KeyBank is committed to building a team of engaged employees to help clients and shareholders achieve financial wellness. Headquartered in Cleveland, Ohio, KeyCorp is one of the nation’s largest financial services companies. Key Companies provide investment management, retail and commercial banking, consumer finance and investment banking products to individuals and companies throughout the United States and, for certain businesses, internationally.
Using Robotic Process Automation (RPA) was a way for KeyBank to solve a full spectrum of business challenges. The goal of automation was to streamline processes to provide a better employee and customer experience. As the financial organization grew, organically and through acquisitions, it needed to make work processes more efficient and keep costs manageable.
After doing an analysis of its operations, KeyBank saw opportunities for business efficiency. Instead of continuing to hire at a high rate, the company decided to implement automation to systemically handle some of the manual workload.
The company created its own internal RPA training program with 175 employees completing the course. This was supplemented with the business analyst class and other learnings offered by Automation Anywhere University.
A major focus was on loans processing and account reconciliation
• Onboarding and funding functions for payment protection program (PPP)
• COVID-19 related loss mitigation processes
• 40K Documents processed with IQ Bot
• 100 Different systems and websites using automation
• 97 Unattended bots in production
• 9 Years of work completed in two weeks
• 9 Bots created for COVID-19 response
• $5M Run rate savings in 2020
Using a blend of RPA and IQ Bot, one of the manual processes automated was for mortgage quality checks. Before a customer gets funded, underwriters must first approve a loan. Data on appraisal documents and flood certifications produced by a third-party had to be completely accurate. The mortgage team was responsible for 100% of the time-consuming, manual reviews every day. The volume quality checks required fluctuated from a few dozen to more than 100 loans during the peak seasons of spring and summer.
“Overnight, we had to figure out a way to respond to increasing call volumes and staff for large amounts of work for
processes that didn’t exist before COVID-19. One of our success stories was in loan origination where we created nine bots that were able to do nine years’ worth of work in just two weeks.” — Dominic Cugini, Divisional CIO of Service Digitization
Partnering with Automation Anywhere, KeyBank came up with an automated workflow for end-to-end processing. Documents are downloaded and turned into PDFs. Those with more than 30 pages are split up. Then, the PDFs are uploaded to IQ Bot which is tailored to recognize the difference between the appraisal and flood certification forms. After this step, a response file in the form of a CSV is sent back for an RPA task bot to do lookups and comparisons. Lastly, an output report of anything that needs to be manually checked is sent to the team. With automation, employees no longer must process these documents and the exception rate has only been 15%. To date, 40,000 documents have been processed through IQ Bot.
“With all the different methods of extracting information from documents we find IQ Bot as one of the best values for the buck in the semi-structured document space. The validation portal is easy to configure and allows the business user to provide ongoing document learning and manage exceptions.” — Michael Reynolds, Business Technology Senior Manager of Service Digitization
When the COVID-19 pandemic created challenges for teams to work on site, KeyBank needed a way to respond to increasing call volumes and amounts of work with no time to add to staff, the company turned to RPA.
“We’re looking forward to use IQ Bot for additional robust capabilities to take a lot of the legwork out of getting a document prepared for what data needs to be extracted. Using handwriting capabilities and Azure OCR will allow us to be able to expand from just printed PDFs to additional document types.” — Ryan Fish, Automation Manager
One of the biggest success stories was in loan origination where nine bots were created to manage the bank’s loan origination process. The bank did not have the ability to recruit to ideal staffing levels in such a short time, though with automation, it was able to complete nine years’ worth of work in just two weeks.
KeyBank continues scaling its digitization program and plans to use IQ Bot for additional robust capabilities such as classifying and preparing documents for data extraction. Particularly, the company is looking forward to using Azure OCR, which can recognize handwriting and markings to allow it to expand from processing traditional printed PDFs to non-standard documents such as credit reports as well.
Currently the company is taking a big step by conducting a proof of concept (POC) for Enterprise A2019, the world’s only web-based and cloud-native RPA-as-a-Service platform. The POC to investigate attended bots is expected to turn into a pilot next quarter.