Scientist II- Matching, Membership, Maps, Experimentation

About the Team

We are hiring scientists across multiple teams, including delivery logistics (matching), membership, maps and experimentation.

• The Delivery Logistics (matching) science team is responsible for building various large-scale technologies which power our global marketplace. These include designing real-time backend matching & scheduling systems, creating new delightful user experiences via different delivery options with intelligent pricing, and serving predictions such as estimated time to delivery (ETD) used for decision making across the marketplace for various pricing systems.
• The Membership science team is responsible for designing quantitative models to grow the Uber One program which offers benefits spanning across both Rides and Eats. These include causal ML models to determine the best setup of benefits, acquisition funnels and churn preventions, and novel experimentation designs and analyses to understand long-term value of membership.
• The Maps science team is responsible for solving some of the largest, most challenging modeling problems in Uber, including travel time prediction (ETA), traffic prediction, route recommendation, navigation, search & ranking, pickup & dropoff recommendation, location intelligence, trip intelligence. The outputs of these models power the core trip flow experience for all products in the Rides & Eats businesses, and core internal decision systems such as pricing and matching.
• The Experimentation science team is responsible for building the experimentation platform and the underlying measurement models at Uber, providing reliable, trustworthy and agile experimentation and experiment analysis to power business decisions across the entire Uber ecosystem.

• Solve ambiguous, challenging business problems using data-driven approaches including ML, Optimization, Causal Inference.
• Develop and implement statistical / econometric methodologies to improve results validity, power and generalizability.
• Develop data-driven business insights and work with cross-functional customers to find opportunities and recommend prioritization of product, growth, and optimization initiatives.
• Design and analyze experiments, present results that provide actionable recommendations.
• Orient the teams around data-driven product development by driving the creation of logging, metrics, data visualization and diagnostic tools, and experimentation paradigms.
• Define how our teams measure success, by developing metrics, in close partnership with cross functional partners.
• Owning the product development cycle end to end from data and science aspects.

• Ph.D., M.S. or Bachelor's degree in Economics, Statistics, Machine Learning, Operations Research, or other quantitative fields.
• 2+ years of industry experience as an Applied or Data Scientist or equivalent (not required for Ph.D.).
• Background in at least one programming language (eg. Python, R, Java, Ruby, Scala/Spark or Perl)
• Ability to use Python or similar technologies to work efficiently with large data sets and prototype algorithms & models
• Coding and SQL proficiency and ability to develop statistical analysis and algorithm prototyping in Python or R.
• Design experiments and interpret the results to draw detailed and actionable conclusions across a variety of key performance indicators

• Ph.D. in Economics, Statistics, Machine Learning, Operations Research or other quantitative fields.
• Excellent communication and collaboration skills: Able to lead initiatives across multiple product areas and communicate findings with leadership and product teams
• Experience in leading key technical projects and substantially influencing the scope and output of others
• Thought leadership to drive multi-functional projects from concept to production.
• Experience in large-scale marketplace algorithm design, membership / consumer growth, or experimentation platform.
• Experience with exploratory data analysis, statistical analysis and model development
• Strong programming skills to prototype models in at least one of Python (preferably), R, Java, Go, Scala