Micro Free Course — 100% Free Learning
All lessons in this module are free to learn. Sign in with Google to save your progress.
Foundations: Why Models Matter
OLTP vs OLAP, normalization 101, the cost of bad models, and the layered architecture (source → staging → marts) every modern warehouse uses. The 'why' before the 'how'.
Module Content
Why Data Modeling Exists — OLTP vs OLAP
One database can't serve both transactions and analytics well. This is why data warehouses exist. FIND_VIDEO: search 'OLTP vs OLAP data warehouse' — recommended channel: Kahan Data Solutions / Seattle Data Guy. Aim for 11 min or under.
Quiz: The Two Worlds and Why They Need Different Schemas
OLTP and OLAP optimize for opposite things. Trying to use one for the other is the source of most warehouse pain.
Database Normalization — 1NF Through 5NF (and When to Stop)
The three normal forms in plain English, with the warehousing context most tutorials miss. FIND_VIDEO: search 'normalization 1NF 2NF 3NF tutorial' — recommended channel: Decomplexify. Aim for 12 min or under.
Quiz: When to Normalize and When to Stop
The textbook says 3NF. The warehouse says 'usually denormalize'. Both are right — for different reasons.
Enterprise Modeling Paradigms — Kimball, Inmon, Data Vault & OBT
What goes wrong when you don't model: slow dashboards, inconsistent metrics, and a 6-month refactor. FIND_VIDEO: search 'data modeling mistakes warehouse postmortem' — recommended channel: Seattle Data Guy / Kahan Data Solutions. Aim for 12 min or under.
Quiz: Spotting Model Smells
Five recurring smells in production warehouses. If you find them, you've found a refactor.
Layered Lakehouse Architecture — Bronze, Silver, Gold & Modern Marts
The architectural pattern almost every modern warehouse uses, and why each layer earns its keep. FIND_VIDEO: search 'dbt layered architecture staging marts' — recommended channel: dbt Labs / Kahan Data Solutions. Aim for 11 min or under.
Quiz: Layered Architecture in Practice
Three layers, one purpose: keep the messy world separate from the clean dimensional model.