Metrics store to AUTOML Forecasting

M1-Nick

dbt released a new functionality referred to as the semantic layer, which is basically a nicer wrapper around a metric store.This metric store makes it really easy to develop single source of truth metrics across the organization. The metrics store enables data teams to slice the data by various dimension and look at time-series views.

As metric store become common practice it would be really nice to see a Dataiku integration with these frameworks (for example being able to write the metrics syntax from the SQL editior in Dataiku).

Beyond just being able to use the metrics logic and views in dbt, it would be a great application for an analyst / data scientist to be able to have a drop down to select a metric from a select list of predefined metrics, develop a forecasting model and deploy it. The same thing would be very useful for anomoly detection with automatic alerting set-up to you communication channel of choice

Ashley

Thanks for your idea. Your idea meets the criteria for submission, we'll reach out should we require more information.

If you’re reading this post and think the ability to leverage a semantic layer would be a great capability to add to DSS, be sure to kudos the original post! Feel free to leave a comment in the discussion about how this capability would help you or your team.

Take care,
Ashley