Fuzzy match is the way of finding entries of the same name in the text.

It operates not only with the exact matches. Even if the name is misspelled – it will work.

Fuzzy match is vastly used by financial institutions, but creating such algorithm is not an easy task.

It has to be able to manage:

typos.
aliases and duplicates.
spelling errors.
nicknames.
swapped components – without fail.

And it is topped by the cherry of volumes.
Even if you have 100,000 customers, when doing periodic checks on a customer, against, say, a sanctions watch list database of 6 Million records, you are sifting through 0.6 trillion pairs of computations.

Here is where Sutra Management intervened and simplified your business challenges.

With over 15 years of analytics and business advisory experience with leading banks and FIs across the globe, Sutra team realized early, that AI enabled smart algorithms were the only reliable way to go.

Our research and development, culminated in the state-of-the-art, patented fuzzy name search algorithms, that power our Financial Crimes Compliance Suite today.

It soon outperformed market leaders by a wide margin.

In benchmarking tests that were run for Sutra to win the patent, our algorithm was tested against best-of-the-best market leaders and comparison was done both one-on-one as well as in mash-ups.

Here is what the outcomes looked like – Test results decisively proved that our false positive rates were at a 15%, whereas the other market leaders were left behind with 40%.

Superior performing algorithms meant at least 25% reduction in false positives in Forensics and fighting financial crimes.

Resulting in hundreds of thousands dollars savings per year, even for a medium size company.

That’s why prestigious financial institutions rely on Effiya and Sutra for mission critical applications, such as Next Best Offer, Fraud & Financial Crimes Analytics, Anti-Money Laundering & Risk Analytics.

Reach out to our Marketing team for your free trial today.

info@effiya.com
contact@sutra-management.com