Why this is in my collection
From the publisher:
A practical guide to building a cloud-based, pragmatic, and dependable data platform with SQL
Highlights
- dbt is half of model-driven — Zagni's approach shows SQL templated with Jinja and macros generating repetitive transformation code; the generation reflex is there, but the model that should drive it stays implicit in the codebase rather than explicit in a modeling layer.
- Tests as declared expectations — dbt tests attach rules (uniqueness, referential integrity, accepted values) directly to models as declarations, one of the clearest mainstream expressions of rules-as-first-class next to data and code.
- Lineage for free from ref() — because dependencies are declared rather than hard-coded, the dependency graph is derivable metadata; structure declared once yields documentation, ordering, and impact analysis without extra effort.
- A pragmatic platform is a layered platform — the book's staging/refined/delivery layering repeats the classic architecture insight that discipline lives in layer boundaries, not in the warehouse engine underneath.
- Version-controlled analytics engineering — treating transformations as software (git, environments, CI) moved data work onto engineering rails; the cultural shift that makes generated, regenerable pipelines acceptable to teams.
Highlights on this page are generated with the help of AI.