There’s no good time to catch someone using your stolen credit card, but doing so as it is happening sure beats discovering it when you get your bill.

But according to IBM only 16 percent of banks who participated in a recent study indicated that they could predict fraud as it was happening.

And the news gets worse. The study also showed that once the method behind the hack is discovered, it can take as long as four weeks before countermeasures can be deployed.

This morning IBM announced that it has acquired Koblenz, Germany based IRIS Analytics, a supplier of real-time solutions for the control of fraud and risk in electronic payments, to help it customers address the problem.

You Can See What Iris Sees

Iris differs from many fraud detection solutions on the market today, which use algorithms to identify suspicious transactions.

IRIS Analytics, in contrast, allows for human experts to look at the expert-driven rules and traditional predictive models that are being used and to tweak them without going back to the vendor. Not only that, but IRIS also leverages machine learning and artificial intelligence so that it gets smarter, quicker.

While IRIS might not be able to keep all the bad guys away, it may have a better chance at preventing them from completing transactions.

While IBM’s initial placement for IRIS will be within the IBM Counter Fraud Management product line, additional use cases exist elsewhere in the portfolio as well.

Terms of the transaction were not disclosed.