There are new realities for managing data and data-centric workloads across the enterprise in a unified and comprehensive manner:

• Use cases were previously focused on efficiently storing and processing data in batch processes. Now there are increasing needs for integrating the entire data life cycle and for processing in both real time and batch.
• Technology infrastructure formerly demanded the co-location of compute and storage to avoid costly network transfers. Now the needs of high-performance analytics drive a move toward disaggregated compute and storage, where each can be sized and scaled independently.
• From a user experience viewpoint, it used to be acceptable to deploy and run in timeframes of weeks, months, or even quarters. Now the expectation is to be able to spin up services in minutes, give users their own clusters, and get insights quickly.
• From the privacy, security, and governance perspectives, the primary concerns were formerly about network perimeter and physical access controls. Now, with the entire data life cycle being managed, operators need fine-grained authentication and authorization at the workload and data layers.

The emergence of the data platform for end-to-end data management has been one of the most significant developments in the data analytics field.