Taylor Miller is fascinated by the endless amount of data to explore and the story one can tell with it.  

Can you share the most interesting part of your career journey?

I’m not sure it’s the most interesting, but it’s the most career-defining moment. I was in undergrad working toward my degree in math and economics and didn’t really know what I wanted to do after graduation. My advisor connected me with some alumni who were working for a data-driven marketing company. My manager at the time wanted to learn this program called “R.” I had no idea what R was at the time, but as a summer intern I was up for anything. After playing, learning, and working in R, I realized that I wanted to find a career in data science. And since that summer internship, I’ve been working to develop my coding and modeling skills to grow as a data scientist.

There is an endless amount of data to explore, and I think it’s fascinating to pull it together and develop models in a meaningful way to reveal a story that was impossible or extremely difficult to see beforehand.

What excites you most about working with data and AI?

The story you’re able to tell. There is an endless amount of data to explore, and I think it’s fascinating to pull it together and develop models in a meaningful way to reveal a story that was impossible or extremely difficult to see beforehand.

Describe an interesting project that you have worked on.

I was consulting for the Readiness and Resilience program under the Joint Chief of Staff at the DoD, where my team worked to predict the deployment life cycle of a service member. In other words, what was the probability that an individual would be able to 1) deploy, 2) complete their deployment, 3) integrate fully and successfully back into society, and 4) re-deploy. The whole process was super-fascinating from the data collection and cleaning to the model prep and modeling process.

We also were able to explore many different modeling techniques, including my first introduction to ensemble models.

We had an extreme imbalance problem that we had to work through, which really helped solidify the importance of looking at different evaluation metrics when reviewing model results. We also were able to explore many different modeling techniques, including my first introduction to ensemble models.