Reference

Modeling Capstone Track

This capstone track validates production-level modeling skills.

Structure:

  • Bronze: grain + star-model query correctness
  • Silver: quality controls + serving metrics
  • Gold: architecture tradeoffs + migration design

Scoring:

  • each capstone is graded on correctness, explainability, and production safety.

How Grading Works #

Each capstone level is graded in two parts:

  • Auto-graded SQL drill: every level (including the warm-up below) carries hidden checks worth 8 points (result correctness, coverage or metric integrity, and an output-column contract) with a pass threshold of 6 of 8.
  • Narrative deliverables: grain statements, quality plans, and migration designs are self-assessed against the weighted rubric printed on each capstone page; Bronze, Silver, and Gold each state their own criteria.

Naming note: Bronze/Silver/Gold here are certification tiers for your skills — not the medallion architecture's bronze/silver/gold data layers, which are a separate concept explained in Medallion Architecture.

Warm-up Drill #

Loading SQL editor...
Tip

Use the capstone assignments like production simulations. Include assumptions, failure risks, and rollout plan in your answers.