Kobi Abayomi

Affiliation: Betaside Recordings
Education: Bachelor’s, Georgia Institute of Technology; PhD, Columbia University

Kobi Abayomi is head of science for Gumbel Demand Acceleration—a Software as a Service (SaaS) company for digital media—and Betaside Recordings, both data science startups at the intersection of demand and supply curves for musicians, artists, and music listeners. He also holds an appointment at Seton Hall University, where he teaches in the undergraduate and master’s in data science program. During the spring 2024 semester, he will introduce a class in data science for digital media in which he will cover methods from the past 10 years in data science—from statistics to generative AI and machine learning. Additionally, he is on the advisory boards of Modal-AI, Barnes & Noble Education, and the Ivan Allen College at Georgia Tech.

Previously, Abayomi was the founding senior vice president of data science at Warner Music Group and director of monetization at Warner Media. He was also a professor at Georgia Tech, the Universidad de Cuenca, and Binghamton University.

Abayomi is a native New Yorker and Manhattanite in particular. He was born at the formerly named St. Luke’s Women’s Hospital on 114th and Amsterdam and his family lived on the east side and then the west side of Morningside Park. He earned his PhD from Columbia University after receiving a bachelor’s at Georgia Tech, so more than half his life was spent between 110th and 120th on Manhattan’s Upper West Side.

Abayomi’s father was a psychologist, and Abayomi remembers him complaining about having to pay someone $300 to do his statistics for his dissertation—that is his first memory of statistics. In college, he wanted to be a physicist but switched majors after taking mathematical statistics (out of the old Introduction to Mathematical Statistics textbook by Robert Hogg, Allen Craig, and Joseph McKean). He says he only got a B but was hooked. Next, he took probability from the future with Otis Jennings (then in grad school at Georgia Tech). Two other memorable operations research/statistics professors were Dave Goldsman and Victoria Chen.

At Columbia, Abayomi worked with Andrew Gelman for a while after asking him for assistance with a paper about measuring environmental sustainability. He also received an Integrative Graduate Education and Research Traineeship for work with Gelman and Upmanu Lall in environmental engineering. He then finished his dissertation with Victor de la Peña on topics that have become relevant to machine learning, did a postdoc at Duke and the Statistical and Applied Mathematical Sciences Institute, and completed a Vertical Integration of Research and Education in the Mathematical Sciences visit with Susan Holmes at Stanford.

Abayomi considers himself an applied methodologist because he has worked on many applied problems in which he has tried to create or derive methodology particular to each. As an academic, he worked on problems from environmental sustainability to econometrics. Since he has been in the corporate space, he has mainly worked on audience segmentation and behavior.

Currently, Abayomi is working on the science behind music listening demand and music creation and curation. He says it is beautiful work on both sides—sounds and affinities. He has found it to be about sufficient representations of a specific neuro-psychological process. This past summer, he and Yifeng Yu—a student in the music information program at Georgia Tech—discovered a precise way to predict demand for a song from just the way it sounds. They presented the work at the Mathematics Association of America MathFest in Tampa and hope to get their paper published soon. That is his most recent proudest moment.