Data warehouses are great for supporting standard reporting and dashboard applications. But they are unwieldy if business users ask questions that weren’t modeled in the data warehouse to begin with. The time, money, and resources required to maintain a data warehouse to support ad hoc queries and exploratory analysis is hard to justify.

Consequently, organizations are exploring new technologies, techniques, and tools that minimize the modeling required to modify data warehouses and support new analytical use cases. Many have implemented big data platforms running on Hadoop and Spark. Besides offering higher levels of scalability, this approach eliminates the need to model data before loading it, accelerating deployments and speeding insights.

In addition, organizations are adopting a variety of in-memory solutions designed to speed query performance and work in concert with disk-based solutions. They are now exploring new data mapping algorithms that automate model design and convert complex operational data into queryable objects and views.

Together, these new technologies eliminate the heavy lifting required to implement traditional data warehouses, speeding deployments and accelerating insights. They reshape the definition of a data warehouse while making good on its promise to create data-driven organizations.