“Data science in the real world is complicated, because the real world is complicated…Practitioners should be prepared for the setbacks—it’s not a clear and direct path.” — Priya Medberry, Class of 2019
Priya Medberry has applied her data science skills at The Walt Disney Company for more than two years. The 2019 alumna of the M.S. in Data Science program at Columbia University and her colleagues conduct in-depth user engagement analyses across digital streaming platforms. They use a range of machine learning algorithms to optimize and expand audience segmentation models to better understand audience behavior.
As an undergraduate at Columbia, Medberry majored in biomedical engineering and statistics. “I have always liked science, tech, math, the STEM subjects,” she said. “And I’m drawn to the idea of tying this work to real life, to seeing a tangible use for the subjects that I studied in school.”
Medberry witnessed the effectiveness of statistical models firsthand while “playing around” to predict sports outcomes. In 2013, the avid basketball fan wanted to prove that Kevin Durant would be NBA MVP, but her models predicted that LeBron James would win, and he did. “I saw how statistics could be objective—they didn’t understand the data, but they were able to predict the outcome better than I could.”
After completing her undergraduate degree in 2017, Medberry chose to stay at Columbia for the M.S. in Data Science program to learn from people with experience in the field, including Andreas Mueller, who is one of core developers of scikit-learn and a former Data Science Institute (DSI) associate research scientist. She also completed an internship at NBCUniversal to lean into her interest in the entertainment space.
“I feel that the program prepared me well in terms of the depth and breadth of knowledge that you need to enter the industry,” she said. “The applied [machine learning] class, for example, is one that I return to often and have found to be really relevant to my work.”
Part of the challenge of her work—and something all data scientists need to be mindful of, according to Medberry—is how cleaning data can impact outcomes. “You want the data to guide you, but you also have to know where to look. You need to understand the data, and also be cognizant of your own level of influence on the analytical process. And you need to be prepared that sometimes the data will prove you wrong.”
Because of its history and reach, Medberry believes The Walt Disney Company is uniquely positioned to set and maintain high standards of data ethics and integrity. “The data landscape is constantly evolving, and anyone that works with data has a responsibility to maintain the most stringent standards,” she said. “I appreciated and respected that at DSI, and I’m fortunate to have found those similar values at Disney. That approach and sense of responsibility is invaluable.”
— Karina Alexanyan, Ph.D.