How to consume PowerBI dataflow output in PowerApps?

[thepuzzlemaster]

Previously all my PowerBI data transformations and output were done through a datamart. This was great as I could consume the dataset in both PowerBI and also in my PowerPlatform flow, which read the data, and then pushed content to a postgres db.

My datamart failed spectacularly for the 2nd time 2 weeks ago, and I cannot bring myself to use datamarts for a third time.

So I rebuilt all of my transformations using a dataflow, and just use a directQuery connection in my PowerBI reports to access the shared data. That side of things is all working great.

However, I cannot find a good solution for accessing my dataflow output in my PowerPlatform flow. All it can do with a dataflow, is refresh the data. But it can't seem to read the data.

I'm not opposed to having my dataflow output stored somewhere else, but I can't find any examples of how to do that other than ADLS2. It's currently hooked up to ADLS2 and appears to store the data there too, but only as csv snapshots. I'm admittedly not well versed in the azure ecosystem or ADLS2, so I may just be missing something here. But when I look to connect to ADLS2 from the PowerPlatform flow, it asks for an individual filename, which I don't see how that can work fro me, since the snapshot filenames change each time a new snapshot is created. Additionally, I am concerned about how it would even parse that csv data.

Am I just missing something simple here?
Is there some other dataflow output I should be utilizing for this? 

I've mostly worked within the PowerBI ecosystem (and still very new there too), so I'm very inexperienced on the PowerPlatform side.

In case it's helpful, ultimately my goal is to read some data from a PostgresDB. Transform and filter that data, and then do a writeback to a different PostgresDB on a different server. I'd like to have this happen every 10 days. Again, this was working fine with my datamart, but I've run into a bit of a brick wall when trying to do it now that my data is using a dataflow.

[thepuzzlemaster]

Okay. I think I found a bit of a workaround solution here. I didn't realize there was a difference between PowerBI dataflows and PowerAutomate dataflows. 

I was able to upgrade my PowerAutomate environment so that it supports DataVerse tables. And then I created a new PowerAutomate dataflow which was linked to my existing PowerBI dataflow. This makes it so that I don't need to do any additional transforms on the data. All of my transforms all live in my PowerBI dataflow. But my PowerAutomate dataflow now outputs the data I eventually want to push to my PostgresDB to a DataVerse table.

Then I created a new PowerAutomate flow which reads from the DataVerse table, rather than my old unusable datamart.

It does add one additional point of failure, as now I need an additional dataflow to run, before my writeback flow will function correctly, but at least I am unblocked.

Still very open to other suggestions if anyone has any ideas for how to improve things.

I'll also add that I recognize that I could just move ALL of my transforms to the PowerAutomate dataflow and then use the DataVerse as my shared data storage, rather than connecting directly to my PowerBI dataflow from my PowerBI reports. I don't have a good understanding of the pros/cons of both of those options. But it would cut down on my number of dataflows if I consolidated them.