Asia's Source for Enterprise Network Knowledge

Sunday, April 23rd, 2017

Storage

Big data is now all about two-speed information architecture

IBM, Box to provide enterprises with option to store data regionally in Europe and Asia

Data and analytics capabilities have made a leap forward in recent years, according to TIBCO Asia head of pre-sales, San Zaw.

For Zaw, with the volume and diversity of available data skyrocketing, computational power and storage improving, and algorithms becoming more sophisticated, businesses need to look at big data differently to succeed.

Zaw mentioned that the volume of data has increased exponentially, as a result of sensor data, and IoT data from wearables, beacons, and other technology.

“We collect information on people, data, and technology but the most important question to ask is why we do that? We do that so customers can achieve their digital journey. And in their desire to go digital, they want us to solve hard problems.

“It’s dramatic. The customer journey that banks and telcos are tracking, retailers are tracking, and Google itself is tracking, for example, shows that a lot of business data is generated,” he mentioned.

As a result, according to Zaw, customers want two things: integrating everything instead of just a few things and solving realtime problems.

“They want to interconnect everything. That includes people, workflows, and processes, APIs, data, and systems. After they get all that data, they don’t just want to sit on it. They want to visualise it on the spot, they want to discover data, not just dashboards; they want to do predictive and prescriptive analytics, and automate that to solve problems.”

Zaw also stresses the importance of two-speed information architecture in IT, which uses visual analytics and predictive analytics.

“Two-speed information architecture is what customers are asking for today. The old model of collecting data into a mass data lake, analysing, and then acting doesn’t quite cut it anymore. They want to be able to still keep the raw data, but act on it instead of mapping it out. To do that, you need this two-speed information architecture,” he said.

According to Zaw, the process all starts with an event.

“All data begins its life as a real-time event. The world is a series of events. Some of them turn into data. That data is usually stored somewhere in a data lake and analysed historically. But insights are perishable, and there might not be enough time to act.

That’s where the two-speed information architecture comes to play,” he said.

The most important and difficult part is what Zaw calls wrangling – cleaning the data and finding the features in it because it is all very time consuming. Then, it needs to be analysed and modelled for prediction.

“Most customers are at the insights phase – they get insights and reports but now, they need to use it to solve problems. You need a model that looks at the data, machine learning, and quick deployment into a business.

“This two-speed architecture is not just relevant for information, it is also relevant in the general IT landscape as well. We’re seeing more customers putting emphasis on lightweight, line of business technology that is towards customer facing. The old rules don’t cut it anymore,” he added.