Bringing cloud computing to data analytics

Data analytics are increasingly important to businesses. It is therefore important that students in the Gies College of Business MBA program have a strong grasp of this area. The ability to make smart business decisions based on an analysis of relevant data has become an essential skill for business professionals.

“Research has shown that evidence-based decision making has greater returns than gut instinct based,” said Business Administration Associate Professor Ramanath Subramanyam, who teaches data analytics courses in the Gies Business MBA program. “Firms that adopt data-driven decision making have output and productivity that is 5-6% higher than what would be expected given their other investments,” he said, based on a recent study. “A student of today has to be very much up to speed in understanding and applying data analytics regularly and wisely.”Ramanath Subramanyam

The data analytics courses require students to learn some basic programming to work through various datasets in a simulated computing environment. Setting up a course that includes such simulations can be a monumental undertaking. There is hardware to spec and purchase, software to be licensed, servers to set up and configure, meticulous calculations of data storage needs to be finalized—all before teaching can begin. In total, there could be many weeks or months involved in creating such a course.

Or that’s the way it used to be.

Faculty at the University of Illinois now have access to Amazon Web Services, which enables them to create and run cloud-based applications and platforms.

Aria Novianto, associate director of research and business technology in Gies Business, worked with Subramanyam and Technology Services at the University of Illinois at Urbana-Champaign to set up the environment for the course.  Setting up the course took much less time than it would have if actual servers were involved. Novianto and Subramanyam estimate that they spent about two weeks’ time preparing the class environment. “I estimate that it was ten percent of what you’d normally need to set up something like this,” said Subramanyam.

The AWS interface is browser-based. So long as students have an up-to-date browser on their computer, they will be able to use the simulator and interface as easily as the instructor can. In other systems, how individual computers interacted with a server could impact the ability of students to run simulations. Previously, interacting with servers could cause difficulties if students’ computers were not set up correctly or lacked needed software. With AWS, said Subramanyam, “if I can run it, I know they can run it, too.”

The ability to scale as needed also made this a very easy way to manage an environment. “Hardware estimation is a very hard challenge,” said Novianto. “Even the biggest companies struggle with that.” But with AWS, scaling up computation needs due to increased student demand was made a nonissue.

Subramanyam was pleased with relative ease with which they could bring high-quality computing environment into the classroom for helping students apply hands-on data analysis techniques and programs. Overall, he said, the experience of incorporating cloud-based computing solutions into the classroom was a smooth and easy one: “What was really a surprise was how effortless it was.”