incorporate puppini's work around the "unified star schema"

[fabrice-etanchaud]

As dbt metadata contains relationships and metrics, one can catch up traps as defined by francesco puppini in his book written with bill inmon.
Then, your project could help build the unified star schema (the famous bridge table).
It's really a challenging project you envision !
I wonder how it could integrate the upcoming dbt metrics server !

[ataft]

Very cool, I'd love to see this grow with more complex scenarios like the unified star schema. Generally, I would want separate this out as to not overcomplicate things for beginners. But then add on a bridge and talk about that in the documentation. I also want to make sure that the "when" and "why" questions are answered. E.g. when and why should I do dimensional modeling instead of one-big-table (OBT), or when do bridge tables makes sense to introduce.

I'll think about this some, but also would love the community to come together on this project. The beauty of this being Github and dbt, rather than in a book, is that everyone can run real examples. I think it's a huge learning opportunity.

[fabrice-etanchaud]

@ataft , I posted this issue just like a message in a bottle... did not think you would answer so quickly ! I totally agree with your way of building on dbt. Time for automation has come, not just technical automation, but also business and reporting. I wonder how data vault could also be incorporated on the way ?

Best regards from French west coast,
Fabrice

[fabrice-etanchaud]

by the way, did you have a look at lightdash and dbt-metabase repositories ?

[ataft]

With regards to datavault, I'm not an expert, but wonder if this package could be useful:
https://github.com/Datavault-UK/dbtvault

I opened a discussion for the topic of lightdash, metabase, and many other existing integrations that I've reviewed:
#7