Module: Setup | Duration: ~5 min | Lesson: 0 of 6

1. What You'll Build

This course is decision-framework focused — less hands-on SQL than the previous two courses, more architectural thinking. The lab is the previous two labs combined: you should have access to the DuckDB-in-Docker sandboxes from Course 2.1 (Vault) and Course 2.2 (OBT). If you do, you have everything you need.

No new tables, no new tooling. The exercises in this course are design exercises — sketches, decision matrices, defenses of architectural choices. Most of your work happens in a markdown notepad, not in SQL.

2. Prerequisites

Course 2.1 (Data Vault 2.0) — completed, or at least lab-set-up.
Course 2.2 (One Big Table) — completed, or at least lab-set-up.
Course 1.1, 1.2, 1.3 (Track 7.1 — Kimball foundations) — completed.

The lab depends on you having internalized the three modeling styles. If you skipped one of the prior tracks, this course will feel ungrounded.

3. Installation

No new installation. Verify your prior labs are accessible:

ls ~/s7-lab/         # Track 7.1 lab — Kimball star schema artifacts
ls ~/s7-vault-lab/   # Course 2.1 lab — Vault sandbox
ls ~/s7-obt-lab/     # Course 2.2 lab — OBT sandbox

If any are missing, return to that course's Lesson 0 and complete its setup.

4. Verify Your Setup

For each lab, confirm DuckDB-in-Docker can still read the seed data:

cd ~/s7-vault-lab
docker run --rm -it -v $(pwd)/seed:/seed datacatering/duckdb:latest -c \
  "SELECT COUNT(*) FROM read_csv('/seed/raw_customers.csv', header=true);"

If that returns a number around 500, you're set.

5. Copy Prompt

I'm starting the "Choosing a Modeling Style" course on data-learning, which is the decision-framework capstone for Track 7.2 (Modeling Alternatives: Data Vault & OBT). This course is more architectural than hands-on — most exercises are design defenses rather than SQL.

What I have: completed Track 7.1 (Kimball foundations) and Course 2.1 (Data Vault), Course 2.2 (OBT). My local labs from those courses are working.

My goal: by the end of the course I should be able to (1) defend a modeling-style choice for a real company, (2) explain the migration cost between styles, and (3) articulate when hybrids are the right answer.

My machine:
- OS: <I will fill in>
- Existing tools (DuckDB-in-Docker, dbt, etc.): <I will fill in>

Walk me through what to expect in the course and how to think about the exercises. If I'm rusty on Kimball, Vault, or OBT specifics, suggest the most efficient way to refresh — without re-doing the full earlier courses.