by Randall Brown
At seven years old, Denise Koessler Gosnell pasted the walls of her family’s RV with colored construction paper adorned with numbered tic-tac-toe grids. It wasn’t your everyday math game.
“I was basically doing matrix addition,” she says. This type of mathematical operation is used to make sense of disparate arrays of information. It can include symbols, letters, or—in Gosnell’s case—colors.
“That’s when my mom was like, ‘I’ve got a different thinker,’” she said.
That afternoon in her family’s Knoxville driveway would be prophetic. These days Gosnell, who describes herself as a solution pattern finder, is a principal product manager for Amazon Neptune. She leads the company’s graph technology efforts— data technology that powers features like recommendations on streaming channels.
In the age of Big Data, graph technology is a way to predict human habits and, in turn, shape decision making.
“This started to become a thing in the early 2000s, when we began seeing a change in everybody’s behavior as they turned to online or digital interactions,” said Gosnell. “That’s when we saw the volume, speed, and shape of data become both enormous and versatile, in a way that we didn’t know how to handle previously.”
The once–aspiring math teacher enrolled at UT with a National Science Foundation fellowship to pursue her doctorate in graph technology, which took a deep dive into machine learning and graph algorithms.
Gosnell is an indisputable leader in the field. In fact, she co-authored the book on it: The Practitioner’s Guide to Graph Data. Now based in Charleston, South Carolina, she most recently served as chief data officer for DataStax, a data management company that provides services to 90 companies in the Fortune 100. She moved to Amazon in September 2021. She ties all of her successes back to her Vol engineering foundation.
“What I studied for my PhD has been at the center of every opportunity I’ve had in tech,” she said.
It was in her doctoral dissertation that Gosnell coined the term social fingerprinting.
“I essentially proved that you are who your friends are, and that understanding data can uniquely identify you,” she said.
Gosnell made her mark in other ways during her studies. She and other women in the EECS master’s and doctoral programs formed the organization Systers—coined from systems analysis—to create mentorship opportunities for women and highlight research at the university.
“The seven of us were tired of being the only women in our classes,” she said. Systers continues to function, and she mentors current members.
Gosnell continues to channel her lifelong enthusiasm for turning up the kind of patterns that now play a profound role in shaping the world.
“I’m so excited to be one of the main leaders in making it easy to use,” Gosnell said.
Likewise, we’re excited to count her among the ranks of our Engineering Vol alumni.