IBM recently announced that it acquired IRIS Analytics, a German firm founded in 2007 that specializes in real-time analytics of payment fraud. Terms of the acquisition, which closed late last year, were not disclosed.
Using Machine Learning to Fight Fraud
IRIS features a real-time analytics engine that uses machine learning to rapidly generate antifraud models. At the same time, it also supports the creation and modification of ad hoc models, which have been proven successful on various payment platforms. The engine helps human analysts detect problems and then act quickly to reduce fraudulent events.
IRIS is already in use by banks and payment processors worldwide. French payment card switching network e-rsb, which is operated by STET, uses IRIS for 5.5 billion annual transactions and has had considerable success with the platform.
“With an average response time of less than five milliseconds per transaction even during peak periods when we are processing over 750 transactions per second, IRIS enables us to detect potential fraud without adding any notable overhead to our service,” noted STET Deputy CEO Pierre Juhen in the official press release announcing the acquisition. “In addition, we are able to respond to newly identified fraud patterns by deploying new countermeasures in a few hours without taking down the system.”
IRIS’s Capabilities Are Impressive
Many other analytics systems require the vendor to institute some sort of process to change things, which is commonly known as the black box model. IRIS takes the alternative: the white box model. This allows the user to change applied rules and evolve the system without vendor input.
IBM has been looking at products in this space for a while now. Bob Griffin, general manager for IBM Safer Planet, told TechCrunch that it went through a build-or-buy discussion. After choosing to acquire a product, IRIS showed up on the radar.
“We began looking at a variety of capabilities, and that’s how we discovered IRIS Analytics,” he said. “[We felt] it was the best positioned and most effective technology offering around real-time payments and fraud prevention.”
One of the reasons this solution stuck out to IBM was its use of machine and cognitive learning. These capabilities help IRIS get better results the more it is used, regardless of whether it’s applied to individuals or institutions.
IBM Continues Antifraud Efforts
IRIS fits in with the other antifraud systems IBM has deployed. Over time, the technology will undoubtedly be integrated with other cognitive learning and security products. In any case, this is not the last the world will be hearing of IRIS.