iHeartRadio is the largest owner of terrestrial radio stations in country, broadcasting continuously to over 850 stations across every major media market. Yet today, radio is just one of many different channels for music discovery. Streaming services and social media are revealing new underground artists and genres to music fans, as well as delivering more information into the hands of industry experts. During this talk, I will discuss how iHeartRadio uses Bayesian techniques to analyze various data sources to deliver the hottest new tracks to over 100 million registered users, while avoiding the most overplayed music. Researchers and analysts who are comfortable with the PyData toolset and interested in applications of Bayesian inference will enjoy this talk.