← Previous · Film 8 of 10
An agent on the team
An AI agent drafts in minutes. It follows rules written in the project, shows the query behind every claim, never weakens a test, and people approve.
Topics › Analytics engineering › In the weeds of data crafting
Written once: one home per fact, and everything else generated
A definition copied into four places drifts in three. Give each fact one home: meaning in the conceptual model, taken from the business glossary where it already defines the term, decisions with why and who in a log beside what they're about, accepted gaps on the model as known limitations, what the build uses in dbt's YAML. Generate the rest, sync it one way, and ship a breaking change as a new version with a date. The next film follows one question from start to end.
Builds on An agent on the team
Watch the film
Press Play, or pick a chapter. Turn on Pause and think to stop for one question at the end of chapters.
Download the videoCaption filesRead the script
Where next?
You've seen one home per fact. Now find the home, and plan the change.
See it in the project
Every card in the film shows a real file from the example dbt project. Open the ones each chapter shows, on GitHub, or run the whole project yourself.
- Four copies
models/core/course/_course__conceptual.yml - What goes where
models/core/course/_course__conceptual.ymlmodels/_shared/_shared__decisions.ymlmodels/core/student/_core_student__models.ymldocs/conventions.md - Written once, shown everywhere
models/core/course/_course__conceptual.ymlscripts/generate/definitions.pymodels/core/course/_course__definitions.mdmodels/core/course/_core_course__models.yml - Diagrams that can't drift
models/core/student/_student__conceptual.mdmodels/core/student/_student__physical.md - One direction
dbt_project.yml - The next version
models/core/student/_core_student__models.ymlmodels/core/student/core_credential_v1.sqlmodels/marts/wallet/mart_wallet__credentials.sqlmodels/core/student/_student__decisions.yml
Think it through
The four Pause and think questions, for a class or a team. Open one to see the answer.
In the series: In the weeds of data crafting
Film 9 of 10.
Next · Film 10 of 10 →
End to end
One question, ten steps, ten commits. Every file in its place, for as long as it's needed.