Richard Buckle
December 17, 2015 · 4 minute read

It’s not good enough today to exclaim, “I just didn’t know!” or worse still, “I didn’t think it applied to me!” And yet, in conversation after conversation with members of the NonStop community they have expressed to me an almost disdain for anything big data related to the point, where from a sense of exasperation, I highlight just how important the intersection of big data and transaction processing has become. If you truly do not want to correlate changing behavior with products and services being offered, then how can you determine whether these products and services provide any value to your customers?

Equally as important, when do you realize that your customers are leaving you and heading to something that is better, more competitively priced or simply, valued as being “cool!”? Without a mechanism to pull pertinent information from the volumes of unstructured data being generated all around you even as you continue to just rely on the arrival of one transaction after another, how is it even conceivable that you will be able to grow your business? It’s been questions like these that I have posed while responding to the ad hoc conversations of late and yet, for as smart as the NonStop community has demonstrated it is, somehow the bigger picture continues to allude it.

Monitoring Social Media Could Have Saved Millions

At its most simplest level of interaction, processes behind transaction processing can be augmented with information on their surroundings as we have become familiar with how to pull relevant information from social media, news feeds and even email and text messages – the wealth of information on sites like Facebook, Twitter and even LinkedIn are just too important to leave untapped. Take for instance, when a couple of years ago it was reported worldwide how thieves executed a “global raid on ATMs with criminal gangs fraudulently pilfering $45 million in two separate attacks.” I wrote of this in a post of May 19, 2013, Are you still sure you are secure?

What led to this successful heist was the crowdsourcing of willing participants prepared to pick up fake counterfeit, open-loop prepaid gift cards. “The New York suspects were U.S. citizens originally from the Dominican Republic who lived in the New York City suburb of Yonkers,” said CNBC. “They were mostly in their 20s. Lynch said they all knew one another and were recruited together, as were cells in other countries.” According to USA Today “teams of so-called cashers allegedly launched carefully timed attacks that caused more than $5 million in criminal losses from more than 4,500 ATMs in about 20 countries. In approximately 10 hours, casher cells in 24 countries conducted approximately 36,000 ATM transactions worldwide, withdrawing an estimated $40 million.”

Essential Data for Mission Critical Operations

If only the banks had systems monitoring social media feeds they would have witnessed the recruiting and initial exchanges between the cashers and as the first attacks on the ATMs began could have shut down their ATM networks and save the majority of funds stolen. These attacks are surely going to escalate in the future as the skills and determination of the bad guys increase – no matter the type of raid launched, somewhere there’s communication being undertaken and there’s always a trail to follow.

For everyone in the NonStop community the years of supporting payments solutions on NonStop have resulted in NonStop systems residing somewhere along the transaction path of almost all ATM as well as nearly all POS and increasingly in the path of every SMS text message traversing the global network. As such, NonStop systems resident software has the potential to thwart fraud before it even gets off the ground. All that’s needed is a solution that allows to better analysis what is being discussed on social media concerning the vulnerabilities of financial institutions. And this is just one example but as security continues to take more of the spotlight it’s important that companies become more proactive in embracing big data and stream analytics.

Streaming Analytics to Thwart Fraud in Real Time

With the launch of NonStop X utilizing blades based on the Intel x86 architecture, the NonStop users have ample power to integrate the output from data stream analytics processing in a timely and meaningful manner. It’s not a complex problem to solve today and the number of deployments of such software is on the upswing. Talk to anyone involved with Striim and you will see how easy it has become to interface the world of data streaming to transaction processing and the benefits are undeniable. If it sounds like I am bullish about the prospects of Striim well, it is because I am bullish.

It is my hope that by the time the NonStop community reconvenes in 2016 at San Jose for the next installment of the NonStop Technical Boot Camp, the conversations I am having today will have receded to being nothing more than a faint memory. “I just didn’t know!” not to mention, “I didn’t think it applied to me!” will have been replaced with, “I am currently addressing the integration of stream analytics with my payments solution!” It’s not rocket science nor is it voodoo magic – although the science behind it is very impressive. As this year comes to an end it’s also my hope that I will be posting more on what the NonStop community is doing to address these integration challenges and I have to believe, for many within the NonStop community, the easiest approach proves to be the path taken with Striim.