Swiss-based IMTF Group uses state-of-the-art digital tools to help banks around the world shift to an intelligence-led approach to financial crime.

At first glance, IMTF’s unobtrusive, stained-white concrete headquarters doesn’t leave the impression that it is sitting at the forefront of global banking technology. Saddled between similar rectangular structures in a mixed-use industrial zone outside Fribourg, it instead looks like an old, repurposed watch factory - not a high-tech company active in 52 countries.

From that unassuming location, it is changing the way the financial sector handles financial crime and compliance. It does so by moving institutions away from «old school» practices such as static alert handling, rigid, error-prone escalation, and then staid, formulaic investigations. Taken together, it all too frequently led to deficient internal advisory services provided by compliance and financial crime professionals.

Instead, IMTF prods them to do what numerous police forces around the world have done in recent decades, which is to adopt an intelligence-led approach. In policing, that means shifting away from reactive, individual incident responses to a broader, risk-based approach using the insights provided by surveillance and agency information-sharing.

The False Positive

IMTF methodology mirrors contemporary enforcement practice. Its case management system filters bank information and standard market tools through its modular components.

In practice, that means looking at a bank’s client account data and things such as Worldcheck’s politically exposed person (PEP) and high-risk database, together with other common adverse news interfaces, and then filtering them through IMTF’s proprietary screening process.

This helps significantly cut the number of a bank’s useless alerts, the bane of legacy money laundering and compliance technology. They are either called false positives - for ones that are not real - or false negatives, for ones that are real but are then discarded and ignored.

Name Screening

When it comes to name-screening, for example, IMTF routines pre-filter potential hits with traditional matching algorithms and then use machine learning and artificial intelligence (AI) in a subsequent, second step. That ensures a higher level of real hits getting through, whereby the bank can subsequently adjust and modulate the level of AI employed to higher and lower levels, further increasing system accuracy as it gains experience with the technology over time.

The machine learning function doesn’t just do plain name matching, a traditional headache for western banks in Asia screening Chinese names, and which proved largely ineffectual on platforms only compatible with the Latin alphabet. IMTF’s AI looks at alerts from the viewpoint of syntax, phonetics, culture, domicile, and semantics. It then passes them through 15 different languages.

For example, it will learn common name order variants of the same individual, Choi Min Sik, and Min-sik Choi. This simple variation could have potentially generated hundreds of useless hits in legacy systems. Or none at all.

Case Manager

The case manager IMTF employs is a simple interface that doesn’t require a specific IT manager to operate. It allows investigators, financial crime, and compliance professionals to explore graphically what has prompted a specific alert. The rules are clear and written in English

Typological Investigations

Their interface also allows for any kind of typological investigation set by compliance or management based on internally defined parameters. They can think up a wide set of scenarios, such as an input to review all clients with assets of more than $10 million US dollars who withdraw multiples of $10,000 a day from two or more ATMs in the same city over a defined period.

Prudent Implementation

In typical Swiss fashion, IMTF limits the number of projects it works on to ensure the quality of what they are delivering, working with clients for extended periods before going live. This prevents them from getting unexpected volumes of alerts when they do. Or, even worse, no one in the bank even uses the new framework.

Both have been the experience of numerous large-scale banks, many in the volume-heavy retail and commercial sectors. When taking a big bang approach with a comprehensive new platform, they were suddenly faced with untold millions of unexpected alerts. These had to be subsequently run down under the close supervision of regulators while both the old and new systems ran in parallel.

In contrast to many of the largest software companies active in the same field, IMTF not only helps clients manage their financial crime risks far more intelligently but also makes sure clients know what they are getting and how to best use it.