Notes from the platforms I build
Practical write-ups on modelling, migrations, and where AI genuinely earns its place in a data platform.
The five things I check before I call a data platform “modern”
Scalability, governance, agility, literacy, quality — the same five things I've checked on every engagement, whether or not agentic AI is anywhere near the project.
2 Mar 2026Engaged by data, not overwhelmed by it: what Lakebase and Genie Spaces change
Most businesses don't have a data shortage — they have a dashboard surplus. Databricks' Lakebase and Genie Spaces attack that problem from two different angles.
10 Feb 2026Kimball or Data Vault: how I actually choose
Both modelling approaches get pitched as universal answers. In practice the right one depends on how many source systems you have and how fast they change.
14 Jan 2026What migrating off an on-premises warehouse actually looks like
The technical migration is rarely the hard part. The hard part is keeping every report correct while you do it.
22 Nov 2025Where LLMs actually help in a data platform (and where they don't)
Not every analytics problem needs a language model. The ones that do tend to share one thing in common: unstructured text.