Hi, I’m Henriette. I’m a researcher focused on the human side of data and machine learning, and manage Spotify’s algorithmic bias effort.
I am particularly interested in the impact that teams’ design, data and metrics decisions have on machine learning outcomes. This includes the feedback loop between products and their users, the gap between people’s rich experiences and machines’ data interpretation, and systems’ wider impact.
I combine quantitative, large-scale data approaches with in-depth, qualitative research to understand both what is happening and why. This also means translating research outcomes into pragmatic business or product direction.
My projects have spanned voice and bot platforms, quality of personalized recommendations, ad moderation, location data interpretation, and human-robot interaction.
Latest publication: Henriette Cramer, Aaron Springer, Jean Garcia-Garthright, Sravana Reddy, Addressing & Assessing Algorithmic Bias in practice, ACM Interactions. Dec 2018. Online.
Next upcoming talk: FAT*2019 tutorial, Algorithmic bias challenges in industry practice. Jan 2019, Atlanta, GA.