Building a scalable experimentation platform
From being a website evaluating method, A/B testing has come a long way. It is a critical piece in digital product development today. We see a wider range of companies which are massively using A/B testing for evaluating updates and new feature releases. As digital product are aiming to reach global audiences and enabling large scale A/B testing for their teams is becoming a whole new challenge.Spotify is an early user of A/B testing for product development purposes. In the past year, they have focused on reevaluating and rebuilding their existing A/B testing platform. The process, digital infrastructure and culture walked hand in hand as their experimentation team was enabling Spotifiers to get feedback from their global user base.
The machine learning meetup group that eventually grew into GAIA, and the GAIA conference, was founded back in 2014 by Karina when working as a machine learning specialist at Dreamler. After that, she left the best coast for Spotify, first in Stockholm and now in New York, where she works as a research scientist. She will talk about her work building an experimentation platform that can be scaled for all the varied needs of Spotify.
