Design and execute experiments to improve forecasting metrics across specific segments
Perform feature engineering on large-scale transactional data (~5M records in BigQuery)
Build user behavior models, including segmentation, propensity, churn, and retention
Evaluate models using strict temporal holdout methodologies and segment-level metrics
Document experiments with clear reasoning — hypotheses, results, and decisions
Collaborate with the ML Engineer responsible for maintaining the architecture and production pipelines

3+ years of hands-on experience in Machine Learning — forecasting, supervised learning on user data, or both
Strong Python skills, including pandas, scikit-learn, XGBoost, statsmodels, and/or Prophet
Advanced SQL skills and experience working with cloud platforms (GCP is a plus)
Strong model evaluation judgment — you understand why random train/test splits are inappropriate for time series data
Experience performing feature engineering from transactional datasets
Ability to work within an existing framework and architecture without unnecessarily reinventing established solutions

Data Scientist

Key skills