1.3 Merging data silos

Let’s imagine the scenario when you use several applications, each of them uses its own database for data persistence. They are from different vendors; they do not communicate or exchange data.

But from your point of view, these applications just do one step in your company workflow. From your angle, the data sitting in these isolated databases (data silos) is related, and you'd love to see them aggregated (read-only), but your current applications do not make that possible.

It's difficult to find an application which does everything you need. You'll end up using several applications anyway. Then you try to make some integrations between them, but if it's hard, then you still have another option: copy the data into another database, aggregate the data and provide an access interface for it.

It's possible to use an archive system for this scenario too, or a document database which provides flexible data schemas.

Finding the right connection points between the silos can be challenging, but if you have some permanent global identifiers, which are not vendor/system specific, then those can be used to make the connections, aggregate the data across the (earlier) isolated silos.