senior-data-scientist
by alirezarezvani
Data & AnalyticsBrief
World-class data science skill.
Description
What This Skill Does
Covers statistical modeling and experiment design for production ML systems. Handles A/B testing with sample sizing and significance testing, feature engineering pipelines with Scikit-learn and XGBoost, cross-validated model evaluation with SHAP explanations, and causal inference via difference-in-differences. Works in Python, R, and SQL.
Bundles experiment design, feature engineering, model evaluation, and causal inference into one skill, so analysis stays statistically rigorous from design through deployment.
When to Use It
- Sizing an A/B test before launching a new checkout flow
- Building a churn prediction model with cross-validated evaluation
- Estimating the causal effect of a policy change on revenue
- Engineering lag and cyclical features from transaction timestamps
- Tracking and comparing model runs across experiments in MLflow
Example Workflow
User asks: Design and analyze an A/B test for a new email subject line with a 10% baseline open rate and a 5% minimum detectable effect
- 1
Calculate required sample size per variant using power analysis with the given baseline rate and MDE
- 2
Output a pre-registration checklist covering primary metric, randomization level, run duration, and sample ratio mismatch check
- 3
Apply Bonferroni correction if the user specifies multiple secondary metrics
- 4
Run analyze_experiment() on collected results and return lift, p-value, and 95% confidence interval
Sample size estimate, experiment checklist, and analysis results with lift and confidence interval interpretation
Installation
clawhub install alirezarezvani/senior-data-scientistRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.