Revolutionizing the Music Industry with Data Science
summary
Claims concrete numbers and runnable tools: 2.5 trillion streaming points, 10 billion daily transactions and 15,000 tracked parameters feed Chartmetric and librosa features into ~700‑feature models. Uplift modeling cuts cost per library add by ~30% in TikTok/Spotify campaigns. Three‑month stream forecasting boosts accuracy ~60% and, via partial dependence plots, reveals spend plateaus that enable a 27% marketing cost reduction. A cloud‑native stack of Stream Lead frontend and FastAPI backend delivers real‑time ML scoring and forecasts into business workflows.
AI-written summary · about
chapters (9)
- 0:00Introduction and Speaker Background
- 0:47Overview of the Music Industry's Challenges
- 1:37AI's Role in the Music Industry
- 4:20Data Sources in the Music Industry
- 5:30AI Applications in Marketing and Strategy
- 7:50Uplift Modeling for Marketing Optimization
- 9:12Stream Forecasting and Marketing Investments
- 11:04Building a Cloud Native AI Platform
- 12:13Conclusion and Final Thoughts
related talks
01h 00mAI Powered Uplift Modeling and Stream Forecasting in Music Industry
29mHow Spotify Does ML At Scale
40mPowering Spotify's Audio Personalization Platform
20mRediscovering music
59mClosing Keynote: Paul Lamere
23mThe unreasonable effectiveness of data for better monetisation
04mMusic, the window into your soul
30mTeam Teardown: Driving Platform Adoption and Success