In the past it would take up to six weeks to implement a rule change in its risk systems. This has been brought down to four to five days, thanks to running risk decisioning in Hadoop. "It is a massive reduction," said Saraf.
"Now with the power of our data it was easier to make it frictionless for some of those existing customers. For brand new customers we were able to identify good customers from bad customers much more effectively and more accurately, he said.
"That has made a huge difference to our business from a shareholder perspective, and more importantly from a customer experience standpoint."
Going forward, Western Union's use of big data analytics is set to expand, with plans to introduce more new data base systems such as MongoDB.
According to Thompson, fears of using open source technologies for enterprise workloads are no longer an issue.
"I have seen open source mature, and we have seen value in finding ways to leverage that in our ecosystem. But it is also a journey - picking the right tool for the right business problem."
Saraf adds: "There is great power in technologies like Hadoop. The challenge for firms like Western Union is that while there are some great open source technologies we have to make sure that make sure they are rock solid because we are dealing with people's transactions and hard-earned money."
"So we are doing our best to be at the cutting edge of technology, but we also have to ensure that we are not introducing risk for our customers. So even with Hadoop we decided to go with a partner like Cloudera in this place, because it has put a lot of structure around Hadoop to make it enterprise class."
Future plans - big data in B2B operations
There are also plans in place to utilise big data analytics across more parts of its business. Saraf says that Western Union will also offer services across its B2B division, which currently serves 100,000 customers.
"The next big area we are going to make a push for is learning from these technologies that are proving huge value in the C2C space and applying them to the B2B, which is a big focus of growth," said Saraf.
"For example, the customers in the B2B area are very sensitive to FX changes, and they want model risk to their business based on changes that are happening.
"We have huge amounts of data relating to how FX changes are happening around the world, and trends within this. It would be easy for us to build a data driven product offering for how to model FX and risk to their business."
Source: Computerworld UK
Sign up for Computerworld eNewsletters.