What resources and process do I need to make a Data Foundation a success?
Companies that approach building a data fabric with a technology only focus will fail. A Data Fabric covers much more than just software, it covers people, governance, process and more, however first and foremost it requires “buy-in”.
The technical part of the Data Fabric, like its name suggests, it not just one technology, it is more of a new layer of abstraction. An abstraction that encapsulates the common data preparation that exists within the most common of data driven use cases today. It includes, but is not limited to, data integration, data lineage, data governance, data catalogs, business glossaries, profiling, quality metrics, master data management, data discovery and more. The concept behind the Data Fabric, is that these components are not stretched across multiple systems, but rather part of a cohesive and consistent platform that is stitched together to form a foundation that delivers ready to use data for
use cases time after time with limited need for adhoc work.
It is not unrealistic that these platforms exist and they should not be seen as “magic”. Rather, we have had enough time in the industry now to realise what are the fundamental parts that most data projects require and that it is now time to stitch those learnings together. The same analogy can be drawn to that of a “database” i.e. companies don’t see the need to acquire a separate “query engine”, “indexing engine”, “backup engine” and more, you instead use or purchase a “database” which has abstracted that all into a nice package.