Modeling Manager: Predicting Customer Behavior

Name: Yimin C.
Job Title: Modeling Manager, Card Operations Decision Management

Applying math to analytics
Prior to Discover, I pursued a degree in Math and became a logistics engineer. At Discover, I started in the personal loans organization in analytics. That’s truly where I built my analytical strength. I eventually moved to portfolio pricing because I wanted to learn more about the card business since it’s a core function at Discover. I worked on portfolio pricing for nearly a year before moving to my current team in portfolio risk modeling. What I like about Discover, as you can tell from my journey here, is that there are always new opportunities to learn about other areas. If you have the interest in something new, a new team or a new technology, you’ll be supported to grow into those areas.

Portfolio risk modeling
My current team is responsible for portfolio risk modeling, which is essentially just predicting customer behavior. We use advanced techniques to better understand our customers. By exploring machine learning modeling methodologies, we continuously improve the accuracy of customer behavior predictions. There’s a ton of new projects on this team with machine learning models, researching new adverse action reason code (AARC) methods, etc. I foresee there will be even more in the future.

Collaboration
While a typical implementation used to take three months or more, one of our latest implementations only took a third of the usual time. Our implementations are so efficient because of how well our teams collaborate together, even globally. Every team member takes ownership of their own task while still prioritizing the overall team’s efficiency. From the Business Technology team, to our Shanghai team, to our offshore tech team teams— we all work closely to beat the timeline and make things happen. Nowhere else have I seen collaboration to this degree.

My authentic self
Outside of work I enjoy traveling, reading and playing softball.