Sr Scientist, Tech

Uber Uber · Consumer · New York, NY · Data Science

This role focuses on designing, executing, and analyzing large-scale experiments for Uber Eats and Uber Rides, performing analytical deep dives into user behavior, conversion funnels, personalization, and retention. The scientist will design frameworks for optimizing business objectives, collaborate with cross-functional teams, and present findings to leadership. Requires experience in experimental design, causal inference, SQL, Python/R, funnel optimization, ML model development, and data processing workflows.

What you'd actually do

  1. Design, execute, and analyze large-scale experiments within the Uber Eats and Uber Rides apps.
  2. Perform analytical deep dives to explore topics including conversion funnels, shopping journeys, personalization & ranking, and user retention.
  3. Design frameworks for optimizing tradeoffs across competing business objectives to maximize platform growth.
  4. Collaborate with Product, Engineering, Design, and other cross-functional partners to understand user behavior and guide future product strategy.
  5. Present findings and recommendations to senior technical and business leaders.

Skills

Required

  • Experimental design and statistical analysis for A/B testing
  • Causal inference and quasi experiment
  • SQL and data modeling
  • Python/R for exploratory data analysis
  • Funnel optimization and user segmentation
  • Machine learning model development
  • Designing and implementing scalable data processing workflows using Apache Spark (PySpark/Scala) or Hadoop)

Nice to have

  • Data visualization and dashboarding tools (e.g., Tableau, Matplotlib)

What the JD emphasized

  • Machine learning model development
  • Designing and implementing scalable data processing workflows using Apache Spark (PySpark/Scala) or Hadoop

Other signals

  • designing and implementing scalable data processing workflows
  • machine learning model development
  • experimental design and statistical analysis for A/B testing
  • causal inference and quasi experiment