Hi experts, hoping I can pose this quetion for the think tank here.
It is not a technical question - but more of a operating model question.
In my org we've an application that is getting re-platformed to a Power Apps based application. The company building the new platform, lets call them NewX, had initially not considered the data team within the org(lets call the team MyX). And for reporting side of the application they suggested Dataflow gen 1 to bring the data and use that for reporting.
When I, as Data Analytics Manager, joined the project, I suggested using Fabric pipelines to land the data, then thruogh medallion architecture make the data available for reporting.
Contractually NewX is supposed to build and maintain reports. Hence I have suggested that MyX will build pipelines and semantic models, NewX can build reports. However, since there might be multiple models, the handshake between MyX & NewX will be complicated. So project & NewX are pushing for building Semantic models and Reports and MyX will build pipelines only.
I've not dealt with this situation where model and reports are built by separate people, let alone separate teams in this case.
Two reasons I want MyX to build models -
for better support model in future for any enhancement and upgrade
during build, model changes may warrant pipeline/upstream changes. Hence those who have access to pipeline should build models. It is true for report builders too, but their need is not as dire and can be waited out for MyX to change model.
What is your opinion on how to manage the builds of models and reports?
Thank you
