Job Description

As a Senior Data Scientist on the AI team, you will lead the formulation, modeling, and evaluation of complex real-world decision problems using statistics, machine learning, and optimization. You will apply rigorous analytical methods to design robust models, evaluate their impact, and provide data-driven guidance for strategic decision-making in a dynamic marketplace.

This role requires strong theoretical foundations, sound statistical judgment, and the ability to reason systematically under uncertainty. Beyond building models, you will be expected to challenge assumptions, quantify trade-offs, and influence decisions through clear reasoning and defensible evidence.

• Modeling and Decision Science: Design, develop, and validate statistical and machine-learning models that support large-scale decision-making and marketplace policy design.
• Problem Formulation: Translate complex business and marketplace challenges into well-defined analytical problems using probabilistic modeling, optimization, and causal inference, while accounting for uncertainty, operational constraints, and behavioral responses.
• AI-Driven Decision-Making: Apply modern AI approaches, including reinforcement learning and LLM-based decision models (e.g., Jev and Clef), to marketplace decisions. Establish rigorous calibration, threshold-setting, and evaluation frameworks to determine when automated decisions are reliable enough for deployment.
• Cross-Functional Collaboration: Partner with product, operations, and business teams to translate strategic objectives into clear analytical questions, modeling frameworks, and actionable recommendations.
• Experimentation and Causal Evaluation: Design and analyze A/B tests, quasi-experiments, uplift models, and response models to measure the impact of decisions, interventions, and policy changes.
• Model Interpretability and Insight Generation: Develop interpretable models and analytical frameworks that explain underlying mechanisms, identify key drivers, and quantify trade-offs rather than simply predicting outcomes.
• Production Readiness and Model Ownership: Collaborate with data and ML engineers to deploy and maintain production-ready models, while retaining ownership of modeling assumptions, evaluation methodologies, and performance criteria.
• Technical Leadership: Serve as a subject-matter expert in statistics and machine learning, critically reviewing analytical approaches, challenging unsupported assumptions, and promoting rigorous modeling practices.
• Metrics and System Monitoring: Define and monitor metrics that capture system performance, user experience, and long-term marketplace health and stability.
• Stakeholder Communication: Communicate complex analytical findings through clear, structured narratives that enable non-technical stakeholders to understand the evidence, trade-offs, and implications of proposed decisions.
• Mentorship: Mentor junior data scientists in statistical reasoning, model development, experimental design, and analytical rigor.

Requirements

• 5+ years of experience as a Data Scientist, or 3+ years with demonstrated end-to-end ownership of high-impact modeling decisions.
• Strong theoretical knowledge of probability and statistics, machine learning, optimization, and causal inference methods (e.g., Difference-in-Differences, synthetic control, and uplift modeling).
• Proficiency in Python for data science and modeling, including NumPy, Pandas, and Scikit-learn. Experience with PyTorch or TensorFlow is a plus.
• Strong SQL skills and experience working with large-scale analytical datasets.
• Hands-on experience in experimentation and statistical evaluation, including metric definition, experiment design, and statistical power analysis.
• Strong analytical and critical-thinking skills, with the ability to reason rigorously under uncertainty and deliver quantitatively sound, data-driven recommendations.
• Excellent communication skills, with the ability to translate complex analytical findings into clear, actionable insights for non-technical stakeholders.

Nice to Have:

• Familiarity with decision models that return calibrated probabilities over bounded outcomes (e.g., Jev, Clef), and with designing human-in-the-loop policies around them.
• Experience with reinforcement learning or sequential decision methods (contextual bandits, offline RL).
• Practical use of LLMs and AI coding assistants (e.g., Claude) in analysis and prototyping, with critical validation of their outputs.