Lillian Carrasquillo, Spotify | Stanford Women in Data Science (WiDS) Conference 2020

on YouTube ↗10m

" summary

Claims diversity reduces unfair bias in models and shows how Spotify’s personalization work actually does it by powering Daily Mix, Discover Weekly and year‑end wraps. Describes a mixed‑methods team structure (data scientist plus user researcher) that runs offline experiments and heavy A/B testing, flags unintended second‑order effects of algorithmic tests, offers a concrete case where a recommendation labeled women in country converts a non‑fan, and pushes cross‑disciplinary education.

AI-written summary · about

" chapters (11)

1. 0:00Introduction and welcome
2. 0:44Role at Spotify's personalization team
3. 1:49Advice for women in data science
4. 2:53Mixed methods insights team
5. 3:57AB testing and algorithmic effects
6. 4:39Popular Spotify features
7. 5:21User experience discovery story
8. 6:43Liberal arts education impact
9. 8:07Bias and interdisciplinary education
10. 8:52Diversity and inclusion at Spotify
11. 10:00Closing remarks

" related talks

• 41mDiversITy mattersJuliette Reinders Folmer, Wouter Groenewold · Codemotion Rome 2016
• 01h 27mPanel - Women in Data ScienceKristianna Chung, Karla Guerra, Kerry Weinberg, Sonali Son, Stacey Ronaghan · Data Con LA 2020
• 04mMusic, the window into your soulChristine Hung · Strata Data Conference 2017 - New York, New York
• 14mA Journey to Data Science: Tools for Equity and Diversity in STEMIleana Fenwick · rstudio::conf(2022)
• 40mPowering Spotify's Audio Personalization PlatformJosh Baer · Strange Loop 2022
• 07mEntrevista - "mujeres en Data Science"Bea Hernández, Bethlem Boronat, Jimena Martínez, Paloma Lucas, Ainhoa Celaya · Women Techmakers Madrid 2018
• 41mWhere are the women?Dora Militaru · CSS-Minsk-JS 2018
• 01h 25mDiversity Is Too Important to Be Left to WomenHannah Dee · ACCU 2022