Replacing Looker with Cube and Superset

Looker popularized the power and usefulness of a semantic layer, but unfortunately decided to keep it (LookML) proprietary. In this event, we’ll show how Superset and Cube provide a viable, modern, and open-source alternative. You’ll see how the open-source data modeling layer of the headless BI platform (Cube) works with the modern data exploration and visualization platform (Superset), so you can deliver dashboards and embedded analytics to your end users.

[@fael81br]

Question: despite the metric being created in cube and the aggregations defined there, should you use a metric from a cube you would need to set the aggregation once again at the metric field during chart creation? If so, what is the advantage of using cube instead of just ingesting data from a dbt curated data warehouse table, in terms of synergy between the tools?

[@misba-ul-zamashaikh8349]

Ideally the metrics defined in cube should go and sit under metrics of dataset in superset along with formats and data types retained.

[@fael81br]

@@misba-ul-zamashaikh8349 I concur that.

[@BrianWoodruff-Jr]

Also, fire the editor. They keep making cuts which then have if feel so jumpy and choppy. I'm guessing the editor is trying to edit out when the speaker waffles on about something, maybe, but it's not done well enough. Perhaps the un-edited cut would have been less distracting for me.

[@BrianWoodruff-Jr]

Video starts at 29:00