Data Scientist

About The Role

The role owns applied data science work on high-stakes problems: forecasting, causal inference, experimentation, and statistical modeling on large-scale behavioral and operational data. Models and analyses built in this role ship to production and drive decisions measured in millions of dollars. This is not a dashboard-only role.

The data scientist partners directly with ML engineers, product managers, and data engineers to take work from exploratory analysis to deployed, monitored systems — and is expected to own outcomes, not just notebooks.

Key Responsibilities

Design and execute statistical analyses, causal inference studies, and controlled experiments (A/B tests) to answer product and business questions with statistical rigor Build, validate, and productionize predictive models — regression, gradient boosting (XGBoost/LightGBM), time-series forecasting, and classical ML — on datasets at terabyte scale Develop data pipelines and feature engineering workflows in Python and SQL, leveraging Spark or Snowflake for large-scale transformation Partner with MLOps and ML engineering teams to deploy models via REST APIs or batch pipelines, including monitoring for drift and performance decay Design experiment frameworks: power analysis, randomization strategy, metric definitions, and post-test statistical inference Translate complex findings into clear, decision-ready narratives for engineering leadership and executive stakeholders Contribute to team standards for reproducibility, code review, and documentation across research and production codebases What We Are Looking For 3–6 years of experience in data science, applied statistics, or machine learning, with at least one model or analysis pipeline taken to production Advanced SQL fluency and strong Python skills (pandas, NumPy, scikit-learn, statsmodels) Hands-on experience with gradient boosting frameworks (XGBoost, LightGBM) and time-series or forecasting methods Practical knowledge of causal inference and experimentation: diff-in-diff, propensity matching, uplift modeling, or A/B test design Experience with big data tooling: Spark (PySpark), Databricks, Snowflake, or equivalent at scale MS or PhD in statistics, computer science, economics, applied math, or a related quantitative field Bonus: experience with causal ML libraries (DoWhy, EconML), stream processing (Kafka, Flink), or publishing/applied research experience