How Microservices Are Transforming Big Data Analytics
One of the primary reasons analytics and business intelligence (BI) software use is so limited is that the applications are too limiting to justify the investment. After the application is acquired and installed, an internal IT organization or external service provider must populate that application with data to be analyzed. Of course, as the business changes there’s a need for integration software that can be employed to refresh existing data or add new data sources. Add it all up and it becomes prohibitively expensive for many organizations.
But with the rise of microservices BI is about to become a lot more of a feature of an application rather than a standalone application. Case in point is Zoomdata, which has developed middleware for integrating multiple types of big data analytics within other applications based on a microservices architecture. Written in Java, Zoomdata on the back end can pull data from multiple sources, including streaming data and static data residing in Hadoop. On the front end, Zoomata presents a series of components that can be dropped into any application.
At present, half of Zoomdata’s business is driven by application developers. The other half comes from internal IT organizations trying to solve a BI integration issue on their own. Because the Zoomdata platform employs microservices, it provides developers with maximum flexibility in terms of the number and types of data sources that cab incorporated, says CEO Justin Langseth. In fact, it’s that very flexibility that drove both SAP and Informatica to separately partner with Zoomdata.
The marriage of BI and analytics software with microservices should go a long way toward making it possible to embed BI and analytics modules into every application. There will remain a need for dedicated BI and analytics applications. But use of those applications will be confined largely to analysts. The average end user wants to be able to simply drill down into a data set within the context of an application without having to wait for the IT department to develop a new schema to support a specific type of query. Without the ability to drill down into data in real time, most end users can’t even remember why they asked the question in the first place, by the time the IT organization comes back with a report weeks or months later.
Zoomdata effectively eliminates the need for any intervention on the part of the IT organization. That alone saves money by not requiring IT organizations to dedicate staffing resources to manually integrate various types of data sources.
In the future, Langseth believes much more of the integration process will be driven by machine-learning algorithms that not only will automate tasks, but also serve up recommendations. Most of those recommendations more than likely will be delivered via a conversational interface such as Apple Siri or Amazon Alexa, he says.
Chances are high that microservices enabled by containers soon will be transforming the entire BI and analytics application experience. Most end users, of course, won’t know how that transformation has been accomplished. But they sure will appreciate the results.
