Organizations have been aspiring for a long time to achieve a unified data model that can represent every entity in an enterprise. This has been a challenge due to various reasons, some of which have been listed here:

• An entity may have multiple representations across the enterprise. Hence there may not exist a single and complete model for an entity.
• Different enterprise applications may be processing the entities based on specific business objectives, which may or may not align with expected enterprise processes.
• Different applications may have different access patterns and storage structures for every entity.

These issues have been bothering enterprises for a long time; limiting standardization of business processes, service definition and their vocabulary.

In Data Lake perspective, we are looking at the problem the other way around. Bringing Data Lake would mean implicitly achieving a unified data model to a good extent without really impacting the business applications, which are good at solving very specific business problems. A Data Lake may represent an entity to its fullest based on the information captured from various systems that owns this data.

With entities being represented with much better and complete details, Data Lakes do present a lot of opportunities to the enterprise to handle and manage data in a way that can help the enterprise grow and derive business insights to achieve enterprise goals. An interesting article by Martin Fowler is worth mentioning here, as he summarizes some of the key aspects around Data Lake in an enterprise at the following link: https://martinfowler.com/bliki/DataLake.html.