All projects

data

A warehouse schema that answers questions fast

Model facts and dimensions, then watch a slow query become instant.

intermediateSQL~24h

Why it matters

Analytics engineering is a growing role and star schemas are its core skill. It also teaches you why normalising everything is the wrong default for reads.

What you build

  • Fact and dimension tables
  • Slowly changing dimensions handled deliberately
  • Queries measured before and after
  • Documentation of every grain

If you want more

  • Add incremental fact loading
  • Add a semantic layer

Build it 5 steps

  • One row equals what, exactly. Get this wrong and everything downstream is wrong.

  • Build the dimension tables

Material for the whole build

Shows you can design for the read pattern, which is what separates an analytics schema from a copied application one.

Get a plan built around projects like this