AI Analytics for Builders

Your VP Slacks you: "Checkout conversion dropped. Can you look into it?" Most people open a dashboard. You'll learn to ask the right questions first — turning vague requests into sharp, decision-forcing analysis.

You'll build: A Problem Brief with 5 ranked analytical questions that each pass a quality checklist

Support tickets spiked. You investigate the same question three ways — in Claude Code, Hex, and Querio — and learn when each tool fits. By week's end, you've run a complete analysis and published a shareable app.

You'll build: A working analytical environment + your first published end-to-end analysis

Checkout conversion dropped 20%. Three teams define "active user" differently. Before you can investigate, you need to align on what you're measuring — then systematically decompose the drop to the specific segment and cause.

You'll build: A metric spec with guardrails + a root cause analysis memo

Power users have 2x retention — but is that cause or correlation? You'll draw causal DAGs, design an A/B test for a checkout redesign, interpret mixed results, and learn what to do when you can't run an experiment at all.

You'll build: An experiment brief OR a causal analysis design (two equal paths)

Great analysis gets ignored when you bury the insight on slide 23. You'll size a $900K opportunity, stress-test your assumptions, and build a 3-slide executive readout that gets the decision made.

You'll build: A 3-slide executive readout + opportunity sizing model + final portfolio package