Within the data world, I am especially confused by usage-based pricing _and_ the push towards data mesh. In a data mesh you now have hundreds of data analysts i

Within the data world, I am especially confused by usage-based pricing _and_ the push towards data mesh. In a data mesh you now have hundreds of data analysts in your company… all needing access (and driving consumption) to the data lake and all the tooling around it like on demand spark clusters, notebooks, lakehouses, dbts, dashboarding, catalogs, governance and so on… it gets really expensive to run what 90% of the times is good old sql. Most importantly, none of the vendors are going to help you optimize resources. You got a standard, highly predictable, BI job each Monday? Vendors will push you to use everything they got, while maybe an on prem server is all you need and it will be cheaper.