Data Scientist - Discovery Mode

Spotify Spotify · Consumer · New York, NY · Music

Data Scientist role focused on evaluating and improving ML models for Spotify's Discovery Mode product, which helps artists find new listeners. The role involves designing experiments, developing evaluation frameworks, and analyzing model performance to drive product decisions and optimize campaigns.

What you'd actually do

  1. Own the analytical function for the Discovery Mode ML squad, driving evaluation and continuous improvement of the models that power measurement and campaign optimization
  2. Partner with ML engineers to develop evaluation frameworks and identify opportunities to improve model performance, reliability, and customer impact
  3. Design and execute rigorous experiments to evaluate model quality, measure outcomes, and guide model development
  4. Conduct deep-dive analyses to assess model performance and translate findings into clear, actionable recommendations for product and business stakeholders
  5. Build, maintain, and evolve dashboards that track model health, customer metrics, and program performance

Skills

Required

  • Python
  • SQL
  • BigQuery
  • dbt
  • data storytelling
  • experiment design
  • A/B testing
  • statistical analysis
  • machine learning model evaluation
  • customer outcome measurement
  • KPI definition
  • product decision making

Nice to have

  • advertising measurement
  • recommendation systems
  • causal inference

What the JD emphasized

  • 4+ years of experience
  • experience measuring customer outcomes
  • experience evaluating machine learning model performance
  • strong data science fundamentals

Other signals

  • evaluating and improving models
  • designing experiments
  • developing evaluation frameworks
  • building analytical foundations
  • measuring customer outcomes
  • connecting analytical insights to product decisions
  • evaluating machine learning model performance