Filene Research Institute Engages Conversations on Finance Technology

Past Tides
May 1, 2018 By Jackie Connor

Filene Research Institute, a credit union and consumer finance think tank based in Madison, Wisconsin, featured a one-day research colloquium, “Oh Shoot! What Now? Technology in a Time of Crisis,” to a packed audience of credit union leadership and staff at the Cove @ UCI in April. Moderated by Richard Swart, visiting researcher at UCI and Filene fellow, the event explored how technology, namely artificial intelligence (AI), can help credit unions and its members prevent and/or manage potential crises situations, with a special session focused on previous natural disasters.

“Today’s AI is more like a baby encountering the world for the first time,” said Bill Maurer, dean, School of Social Sciences, Ph.D., anthropology and the event’s opening presenter. “New AI is leveraging to make assumptions about us to help us navigate the world.”

Maurer, who explained previous AI efforts from decades past operating much like a deductive-thinking chess player, elaborated on this current inductive state of AI being first incorporated into fraud protection and how technology continues to help solve for these ‘oh shoot’ moments. He described fraud protection through a consumer’s unexpected credit card charge, which is typically followed by a text message.

“But there is a lot more to this than just fraud protection–we also think a lot about opportunity protection,” said Maurer. “Having this data allows you to do predictive analytics to bring your members to where they potentially want to be.”

Fifty-eight percent of millennials express concern about collecting too much personal information, but 61 percent would trade some of it for a discount, according to a LexisNexis 2016 study. From this, Maurer explained precarious economics creates a need for income volatility solutions.
Maurer described everyday decision making and other engagements with technology at times does not fit into the way technology is directing us, which is also being studied at UCI.

UCI Panel Discussion-Machine Learning and Deep Data

Later in the morning, the crowd was divided into two different sessions, one focused on deep data and machine learning driven by a panel featuring experts in the fields of economics, financial technology (fintech) and policy/regulation at the Beach, the one of the main event spaces at the Cove.
Sameer Singh, UCI professor of computer science, Ph.D., computer science, opened up the panel discussion about the scalability, accuracy and cost-effectiveness of rising new technologies.

“We have more access to not only who has applied for loan, but also for finance-specific demands,” said Singh. “There are easy ways to put machine learning systems into your pipelines to make them more efficient and gather more data.”

Machine learning has become imperative to companies interested in incorporating big data to their predictive processes.

“The complexities of these algorithms are becoming inscrutable,” said Singh. “We don’t always know why they’re doing what they’re doing.”
Singh’s UCI research group focuses on providing interpretable explanations for algorithm complexities to provide more insight into machine learning with the goal of being able to explain any existing or future algorithm.

Matthew Harding, UCI professor of economics, Ph.D., philosophy and economics, encouraged the audience to use discretion when incorporating machine learning and big data. With big data, Harding recommended not only to consider volume, velocity and variety, but he also encouraged the audience to think about the value of big data, and challenged the audience to question the bottom line.

“One of the things we are able to do today that we’ve never done before is take big data sets and relate them,” said Harding. “The value that big data is really giving us is when we have access to data sets that allow us to link many different layers to achieve that kind of depth.”

In Harding’s presentation, big data was highlighted as a way to explore the relationships between large data sets, as well as predict future outcomes and explain how the world will change. His deep data lab is currently working on the credit data of three million people in the United States to better understand the data relationships.

Chandra Middleton, JD, UCI Ph.D. candidate, concluded the panel discussion and talked about how credit unions have brought in AI and machine learning into their process in addition to how credit unions can participate in law and regulation, specifically on the topic of fraud and financial crime identification and prevention.

Middleton talked about companies like Frogster Interactive Pictures, a publisher of online games, which uses biometric authentication to detect and prevent fraud and financial crime identification and prevention, as well as simplify the user experience. Additionally, she highlighted companies who utilize predictive analytics to create personalized financial wellness solutions for clients.

“If we do not adopt AI-based products and processes, we risk losing market share specifically because our credit union communication channels won’t be as cutting edge as banks,” said Middleton.

Lessons Learned from a Year of Natural Disasters

The concurrent session located in the Cove’s screening room featured three speakers from credit unions across the nation who discussed lessons learned from a year of natural disasters, including Hurricane Maria and the California wildfires of 2017.

Jeff Johnson, chief information officer, Baxter Credit Union shared a heartfelt promotion video of his team’s efforts at their Puerto Rico location post-hurricane Maria and reflected on his big takeaways from the September 2017 natural disaster.

“Assume you have nothing and start from there,” said Johnson. “Your number one entity are your employees—take care of them first and they will take care of the members.”

Tony Hildesheim, senior vice president and chief information officer, Redwood Credit Union, described his company’s response to California’s 2017 wildfires in Sonoma County, where 24 staff members lost their homes. He encouraged the audience to practice communications with staff and to help them embrace the company’s greater purpose.

“Build a purpose greater than financial services,” said Hildesheim. “It’s about people helping each other. The credit union should be the shoulder for the community.”

The colloquium’s second half featured Arun Ross, professor of computer science, Michigan State University, Ph.D., computer science, who discussed how organizations and individuals utilize biometrics and facial recognition as a form of protection.

Fintech Support

The event closed with Chief Business Development Officer and General Counsel at States Title Adrienne Harris, who discussed how regulation changes affect technology to help alleviate disasters and support communities in times of crisis.

“No other industry has been effected more than financial services,” said Harris. “Financial services are the very foundation of our economy, it’s an integral part of every other industry.”

She encouraged the audience to engage fintech companies on the current systematic issues as well as industry regulators to help them figure out how to use fintechs.

“Better financial services technology can be the difference between [a consumer] making it through the month and not,” said Harris. “Fintechs cannot and should not be understated–credit unions are very well positioned to lead the charge.”