Enterprise Data Architecture Trends for 2019

The coming year will be one of big change in enterprise data architecture. Here are the trends you should build into your plans and expectations now.

The world of data is rapidly changing. Data is the prime foundational component of any meaningful corporate initiative. The means to manage the prime asset of data is a key decision point being made continually in competitive organizations. Incorporating new information into this process is required, and tradeoffs must be considered.

A large part of the growth can be traced to integrating the cloud into data architecture-related products. Cloud use has become paramount to corporate efficiencies, and those solutions that embody a solution tightly integrated with the cloud provide the most value.

It is a fascinating, explosive time for data architecture. The two key drivers of the market today are the explosive growth of data science and cloud computing. It is a mix that includes preparing data for artificial intelligence and machine learning, which cannot happen without cloud computing.

However, this is only the beginning of the journey to data engineering as the embodiment of what is to come in enterprise data architecture trends for 2019.

The line between operational and analytics realms in the organization is blurring. Although most personnel in 2019 will still identify their projects with one or the other, many will begin realizing the distinction does not matter to their application. Not only will the hard distinction be removed between major silos, it will need to be removed from intrasystem flow. Analytics needs to flow throughout the ecosystem, and it is the enterprise data ecosystem entity that will frequently be considered in 2019.

Data warehousing is still the face of reporting in most organizations. It is still where we find the most data investment "bang for the buck." However, the ingestion tier is dramatically changing.

Some future processing is going to occur in a cloud storage tier that excludes the data warehouse. The data moving to the data warehouse will be a subset of data lake data, although as long as the cloud storage data lake exists, there will be a strong data warehouse in the mix. Gone are ETL tools for moving this data, and fast languages (such as Python) and the use of Spark will take up the data warehouse load cycles.

The elegant data warehouse architecture is columnar and uses a considerable amount of memory. It's in the cloud for sure, and it utilizes all the benefits of the cloud. Rapid provisioning, elastic scalability, and the separation of compute and storage will be givens for major data warehouse activity in 2019.