© 2019 SourceMedia. All rights reserved.

Fighting fraud in store-brand credit cards gets assist from AI

(Bloomberg) --All shoppers are familiar with the offer of a discount at checkout in exchange for taking out a store credit card. Their wallets may already be stuffed with plastic bearing the names of retailers including Amazon.com, Gap, JC Penney, and Lowe’s, but there’s often one company behind all of them: Synchrony Financial in Stamford, Conn., which is one of the world’s largest issuers of store credit cards.

This summer marks five years since it spun off from its parent company–General Electric Co.–and ever since it’s been building a digital strategy that uses artificial intelligence and machine learning. Synchrony now deploys AI to speed up everything including credit approvals and fraud detection.

Chief Information Officer Carol Juel has led the undertaking. She spoke with Bloomberg Markets about the competitive advantage AI gives the company.

Jenny Surane: You’ve been with Synchrony Financial, including its predecessor GE, for 15 years. How did you get to where you are now? Can you walk us through what led you to pursue a technology career within finance?

Carol Juel: I graduated college and had a degree in classical languages and literature and a minor in math—that’s actually the intersection of computer programming. I worked in computer labs all through college, so when I was coming out of school, I had this interest in computers and computer science. But I was one of those girls who fell under the radar and didn’t take computer science.

I worked for nine-plus years in technology consulting and had an opportunity to see many different financial services companies through that time. Then I really decided I wanted to venture into the industry, and that’s when I joined GE Capital Retail Finance in 2004, which was the predecessor to Synchrony.

High End Data Cables Feed Into Servers

The toughest job I ever had was supporting and helping technology for operations and collections during the financial crisis. It was obviously a time when every financial institution was trying to find efficiencies, and drive costs out, and improve the customer experience. That was a huge learning opportunity around how the business operates. Then in 2011 I became the CIO of the business and obviously later began the IPO journey.

JS: That’s right. In 2014 your company went from being a subsidiary of a massive global conglomerate to its own publicly traded company. I understand you had the daunting task of rebuilding technical systems from scratch. So what were some of the priorities you focused on at the time?

CJ: It was a once-in-a-lifetime opportunity to birth a Fortune 175 company through an IPO out of a big company. As we were positioning the company, technology was core to the future. It’s all the things you might think about—it’s how you design certain pieces in software, services, and cloud; it’s how you rebuild data centers; it’s a lot of infrastructure work.

It’s really [because of] this process that technology is in the DNA of our company. Everything we did—touching all of our partners, all of our customers—had to [be done] in a way that was not at all disruptive. That was the secret sauce. How do you do all of this and still run your business? The analogy I used to say is it’s like driving the school bus down the Parkway—that’s the crazy road near us—in a snowstorm. You have to get the bus where it needs to be and make sure the kids are safe and don’t hit a tree.

There are a lot of challenges in the industry with legacy technology that prohibit you from making certain advancements or they make them more difficult. Those investments we made in 2013 and 2014 are paying off now because we were able to think about the future, and we’re now leveraging those investments.

JS: Which investments in particular do you think have paid off the most?

CJ: It sounds so simple, but we had to build data centers. In building the data centers, we designed them in a way that allowed us to organize our data very specifically. It allowed us to think about a digital-first strategy and lay the groundwork for our private cloud infrastructure. Every bank today has these centers, but very rarely do you actually get to redesign them from the ground up.

JS: Many investors want to know how much companies are investing in AI and machine learning because of the efficiencies and competitive advantages those technologies offer. What are some of the things Synchrony has done to show the market that you’re serious about introducing AI and machine learning technologies into your business?

CJ: Within my organization we have a dedicated AI leader that we put in place almost three years ago. That is a leadership job that reports to me. We [also] have technology leaders on our board of directors. When you get to build a company from the ground up, you get to decide on what’s important to you and what type of guidance you need. We have a technology committee of the board, which is rare in this space when you benchmark that relative to our competitors.

JS: How is AI affecting workflow internally?

CJ: As part of our AI strategy, we’re also using smarter tools known as robotics process automation to automate time-consuming tasks. As a result, Synchrony has reduced process cycle time by 50% with projected savings of more than 10,000 hours of manual work. All this enables our people to focus on more complex problems.

JS: Lowe’s, Gap, and JC Penney are some of your biggest partners, so you know all too well about the dramatic changes happening in the brick-and-mortar retail industry. What sort of technologies are you developing to help your retail partners compete in the mobile and e-commerce space?

CJ: Our cloud and AI strategies allow us to innovate with our partners [and perform operations] much more quickly. Speed is critical. The digital application platform that we’ve rolled out across all of our customers simplifies the consumer experience by allowing them to apply for credit in the merchants’ mobile app. We took [the customer approval process] from many fields down to three. The idea that we can get someone through a credit underwriting process digitally very quickly is critical for our partners and our customers as we continue to see our digital credit application volume increase. Today, 50% of our applications come through the digital channels, so ensuring that the technology we’ve built to support that meets both our merchants’ and our customers’ expectations is critical.

We’re also a data-driven company. That gives us an edge. Frankly the types of data used in our underwriting have grown by five times in the past 20 years, and that’s still expected to grow. We’re constantly exploring internal and external data to better pinpoint what drives the credit-worthiness of a person, drawing deeper data insights from our call centers, with our partners, and from external data such as phone and utility bills and rent payment. Integrating the different types of data into a sophisticated platform helps approve more customers.

JS: Synchrony has also been using AI to tackle fraud. How is that technology allowing you to detect unauthorized charges in ways you couldn’t do so before?

CJ: If you’ve ever experienced fraud in your life, it’s a difficult experience. We’re scaling machine learning models to detect patterns of fraud at twice the speed. By pairing data with real-time machine learning models, we’re able to predict synthetic fraud cases, which helps protect our customers and allows us to continue to adapt to future threats.

If you build models and aggregate [information], you are able to catch fraud sooner. The sooner you can catch fraud on an account, the better the experience. We want to be able to find it before our customers do.

JS: Synchrony is developing quite a few different AI innovations at the moment. What sort of technologies excite you the most?

CJ: I think voice is one of the most exciting spaces for us. I think customers are getting very comfortable using their voice to interact digitally. [Our customers are now] making voice payments through [Amazon’s] Alexa. That’s really the tip of the iceberg in terms of where we think intelligent, automated voice solutions are going to go. The next wave of voice is going to be really compelling. It’s going to be more integrated into people’s homes; it’s going to be more integrated into their cars.

When you think about designing for voice, it requires a lot more AI. The intent engine–which is what drives the response to the consumer–has to be designed in such a way that we all ask the same questions, and we’re intending to get the same answer, but we asked it three different ways. That’s where a lot of work is being done now, and it’s really, really exciting.

Surane is a finance reporter at Bloomberg News in New York.