Turn evidence into decisions
Learn to ask precise questions, retrieve and clean data, reason statistically, communicate findings, and productionize repeatable analysis.
The learning path
Do not treat the resource links as a reading list. Learn enough to produce the outcome, then move forward.
Milestone 1
Translate ambiguous business questions into measurable definitions, then retrieve and manipulate the required data.
CheckpointAnswer a scoped product question with a documented SQL-to-Python workflow.
Milestone 2
Audit provenance, missingness, duplicates, outliers, units, and assumptions while keeping every transformation repeatable.
CheckpointCreate a tested cleaning pipeline with a before-and-after data-quality report.
Milestone 3
Use distributions, estimation, confidence intervals, hypothesis tests, and simulations without overstating evidence.
CheckpointQuantify uncertainty and defend the assumptions behind a statistical conclusion.
Milestone 4
Select truthful encodings, reduce noise, design accessible charts, and connect analysis to a decision rather than a dashboard count.
CheckpointPresent one insight as a concise narrative with a decision-ready visual.
Milestone 5
Build useful baselines, segment behavior, design experiments, analyze causal assumptions, and evaluate practical impact.
CheckpointDesign an experiment and build a baseline model that informs a real decision.
Milestone 6
Turn one-off work into tested pipelines, scheduled transformations, semantic metrics, monitored dashboards, and reliable decisions.
CheckpointProductionize an analysis with automated checks, documentation, and a refresh schedule.
Capstone project
Deliver an end-to-end decision memo backed by a reproducible pipeline, tested analysis, interactive dashboard, and experiment plan.