• Master complex SQL—CTEs, window functions, analytical aggregates—to query millions of rows with sub-second latency.
• Build production-grade features in Pandas & Polars; handle missing data, encoding, scaling, and feature stores.
• Design experiments with proper control/variant logic and power calculations to reduce “false wins.”

• Craft executive-ready dashboards in Plotly and Power BI; apply colour theory and accessibility best-practice.
• Translate raw metrics into narrative arcs that move stakeholders to action.
• Build an interactive self-service analytics app and embed it in a live URL.

• Refresh probability theory, confidence intervals, and hypothesis testing for real-world decision-making.
• Implement linear/logistic regression, regularisation, and model diagnostics in scikit-learn.
• Apply uplift & causal-impact analysis to answer “Did the campaign cause the lift?” not merely correlate.

• Train and tune tree-based, ensemble, and gradient-boosting models that handle tabular business data.
• Use cross-validation, ROC-AUC, and cost-based metrics to choose the best model for the P&L.
• Package notebooks into reproducible MLflow experiments for team hand-off.

• Wrap models in FastAPI, containerise with Docker, and deploy to a cloud function or Kubernetes cluster.
• Automate tests and CI/CD pipelines with GitHub Actions; implement basic model-drift monitoring.
• Document your REST endpoints with OpenAPI for front-end and product teams.

• Work in teams on a real client data-set (retail, energy, or fintech).
• Deliver an end-to-end solution—data pipeline → model → cloud dashboard—hosted on Vercel or AWS.
• Pitch to a panel of hiring managers; receive line-by-line feedback and a shareable reference letter.