Joel Persson — Research Scientist, AI & Economics, Spotify

I am a Research Scientist in the AI & Economics lab at Spotify. My research focuses on econometric and machine learning methods and their application in the digital economy. I am particularly interested in causal inference and statistical decision theory for problems in experimentation, personalization, and policy evaluation, as well as the use of digital technologies in business and public organizations more broadly.

At Spotify, I lead research and development in how to improve the efficiency and effectiveness of experimentation, evaluation, and training of large-scale recommender systems, including how AI can augment traditional approaches. The methods and tools I have developed are used internally for product innovation and run in production on the Spotify Homepage.

I received my PhD from ETH Zurich in 2024, advised by Stefan Feuerriegel and Florian von Wangenheim. During my doctorate, I visited the Operations, Information, and Technology group at Stanford GSB, hosted by Jann Spiess, and interned as Machine Learning Research Scientist at Booking.com. I also contribute to the nonprofit Algorithm Audit.

I hold master's and bachelor's degrees in Statistics and Business & Economics from Lund University, Sweden. Before my PhD, I worked on marketing mix models at GfK (acquired by NielsenIQ) and in performance marketing at Precis Digital.