Hello Is it possible to get data from an already transformed database in Greenplum and just make dashboards over it?

Hi, I'm not familiar with Greenplum, and I didn't find a Singer Tap for it. Does it use a PostgreSQL dialect?

Yep

So Meltano uses Singer Tap to extract data from data sources, which are python connectors we integrate it.

Hum, so Meltano is meant for cases where I necessarily need to pass through all the steps?

You could totally have only modeling on top of your database, as long as we can connect to it.

Nice, thanks

Hum, so Meltano is meant for cases where I necessarily need to pass through all the steps?

Good question @cesar! I agree with @micael_bergeron. As the project grows, it will become more clear a) how you can and b) why you should utilize the web app or the CLI including c) what you can do with each one of the functionalities of the project and d) when to use which, see also https://gitlab.com/meltano/meltano/issues/1149.

@cesar: We use redash for making vizs its a good option incase you can’t make meltano work for you

@cesar don't get me wrong, I'd be delighted to have you use Meltano. I think it really comes down to who is the consumer of the data, and can they write SQL and self-serve. The cool thing about the Meltano UI really is about the Model layer that helps to segment the data and provide an easy query interface.

You'll be able to use Meltano Analyze directly on top of a transformed schema by connecting directly to that schema (instead of the

analytics

one that Meltano uses by default) and adding custom Meltano Models. In order to do so, we'll have to allow Meltano UI to set the connection without running an ELT pipeline. This is a feature we are internally discussing. We'll keep you posted if we decide to support it in one of the following Meltano releases.

In order to do so, we’ll have to allow Meltano UI to set the connection without running an ELT pipeline.

@yannis_roussos this sounds interesting, is there an issue for this?

@yannis_roussos this seems to be a interesting point because I thought from the initial page I thought we can run sqls directly.