Senior Scientist - Tech, Science

• Own experimentation for personalization. Design, run and analyze experiments for ranking, recommendation and agentic features. This includes interleaving, holdouts, switchbacks, and variance reduction (e.g., CUPED/regression adjustment). Use difference-in-differences, synthetic control, instrumental variables, regression discontinuity, or propensity methods, and be clear about the identifying assumptions.

• Close the gap between offline and online evaluation. Build offline evaluation and simulation frameworks (e.g., NDCG/MRR, counterfactual and off-policy evaluation). Diagnose why offline gains don't show up online, including novelty effects, distribution shift and training-serving skew.

• Go deep on the data. Run deep-dive analyses of rider behavior, funnels, and ranking and recommendation outcomes. Include heterogeneous treatment effects and long-term versus short-term tradeoffs. Surface proactive insights that change what the team builds next.

• Define success, guardrail and long-term metrics for personalization surfaces. Make results easy to find and use through dashboards, query templates and debugging tools.

• Shape the roadmap. Turn analysis into clear recommendations for PMs, engineers and applied scientists, and help the team decide which bets to prioritize and which to stop.

• Raise the scientific bar. Review experiment designs and model evaluations, mentor junior scientists, and share best practices across the team. Partner with applied scientists on experiment design for new ML approaches (e.g., listwise ranking, reinforcement learning, evidential deep learning, LLM-based/agentic applications).

• Stay current on advances in causal inference, experimentation and personalization AI/ML.

• M.S. or Bachelor's degree in Statistics, Economics, Computer Science, Mathematics, Operations Research, or another quantitative field.

• 5+ years of industry experience as a Data Scientist, Applied Scientist, Research Scientist, or equivalent, with a track record of owning experimentation and/or causal inference end to end.

• Deep expertise in experiment design and analysis (power, variance reduction, multiple testing, interference/network effects) and in observational causal inference methods.

• Working knowledge of machine learning. You can critique and debug models, understand ranking/recommender approaches and common failure modes, and use ML as an analytical tool.

• Proficiency in Python and SQL, plus PySpark or similar big-data tools. You can build production-quality analysis pipelines.

• Strong business and product sense. You can frame ambiguous problems, take them end to end, and communicate recommendations clearly to technical and non-technical partners

• Ph.D. or M.S. in Statistics, Economics, or a related field with a causal inference or econometrics focus.

• Experience evaluating ranking, recommender, search or feed systems (offline vs. online evaluation, interleaving, counterfactual evaluation).

• Experience with heterogeneous treatment effects, uplift modeling, or long-term effect estimation/surrogate metrics.

• Experience measuring or evaluating LLM-based or agentic applications.

• Hands-on experience with ML frameworks (scikit-learn, PyTorch or TensorFlow), and with good practices such as design documentation, testing and Git.

• History of mentoring and raising the technical bar for a team.

For New York City, NY-based roles: The base salary range for this role is USD $190,000 per year - USD $211,000 per year.

You will be eligible to participate in Uber's bonus program, and may be offered an equity award & other types of comp. All full-time employees are eligible to participate in a 401(k) plan. You will also be eligible for various benefits.