In 2019, CIOs should strive for flexible analytics architecture

Data and analytics are part of everyday survival for companies. But according to a new research note from Gartner Inc., organizations are struggling to manage the data they have let alone establish a plan for the data that’s coming.

“This fruitless trend will continue unless technical professionals act now in preparing for the future demand,” stated the “2019 Planning Guide for Data and Analytics” report, which published earlier this month.

To that end, Gartner analysts identified five trends that CIOs and information architects should pay close attention to in the upcoming year, one of which is how the use of artificial intelligence and machine learning will become useful information management tools.

The foundation of the report is centered around Gartner’s logical data warehouse or LDW, a data management practice born out of the big data era. Rather than build a single, central data repository, Gartner began recommending that companies create a single view of the data through a virtual data layer that connects data repositories scattered around the enterprise and in the cloud.

The analysts caution that there is no single blueprint for establishing such a flexible architecture nor will it happen overnight. Instead, the report states, “the shift will be gradual and incremental — but also inevitable.”

With an LDW in place, Gartner analysts now recommend that CIOs establish an analytics architecture that’s flexible enough to support both traditional as well as newer analytics techniques. The architecture should utilize a plug-and-play framework that can integrate external and internal data sources together and combine the data collection, analysis and delivery of insights into a single discipline.

The ultimate goal of such an architecture is to accelerate digital transformation and support an “analytics everywhere” environment, where analytics can be delivered even at the edge of the enterprise. The detailed report walks CIOs and information architects through all five trends and provides planning suggestions for the coming year.

One trend is that AI and machine learning coupled with an LDW can create a more intelligent way to manage data. The benefits are a two-way street: An LDW can provide AI and machine learning the high quality data needed to produce good results and the infrastructure needed for model deployment. And AI and machine learning can help manage complex workloads, query data efficiently, and size up both the data type and content as it enters the LDW, which the report described as “one of the most exciting areas of the market today.”

To get started in 2019, the analysts recommended that CIOs and information architects look for use cases such as workload management where introducing AI and machine learning could make a dent on performance or where there’s some kind of overlap between the LDW and AI and machine learning such as data quality.

In most enterprises, data quality processing is for data warehouses are separate from AI and machine learning work. CIOs may want to collapse those two efforts together and use “the industrial strength data transformation and quality tools used for the LDW for machine learning,” according to the report.

They also suggested that CIOs take stock of the kinds of tools that are already at their disposal. According to the report, “most commercial database management system software has useful libraries of the most popular ML algorithms. When new analytical requirements can be met by common ML algorithms, these incumbent libraries provide a simple and low-cost means of meeting analytics requirements.”