Another option is data mesh. We mention data mesh as it is an emerging solution to problems with data warehouses and may replace some functions of data warehouses, data lakes and data marts in the future.

It is a data platform architecture that leaves the data at its source, instead of putting the data into a monolithic data warehouse or data lake that handles the loading, transformation and consumption of the data in one place. Data are accessed and used by the part of the business needing it through a layer of interconnecting data pipelines that have universal interoperability, applying the same data standards using automated software extraction tools or API interfaces.

This is a simplified diagram that shows how two different business units are using their own data pipelines to extract the data that they require directly from the same source systems. What they extract could be the same data or an entirely different data set.

How two different business units can extract data from the same sources.