The idetect R&D Team rolled out new name matching features in idetect v6 for Compliance and Anti-Money Laundering purposes that make use of Artificial Intelligence and Machine Learning technology, effectively stepping up competition with historic players in the domain.
The R&D team has created more effective models that reduce false positive alerts by an impressive order of magnitude. These new features help compliance and sanction investigators to drastically improve alert performance and reduce noise. This kind of technology has practical applications in other domains as well, such as law enforcement, defence, census and record linkage.
In this context, idetect’s latest features have been compared with a world-class solution used by the U.S. Dept. of Defense, U.S. Customs and Airbnb… and the results are simply stunning: idetect was ahead 85% vs. 77% in terms of accuracy, and with a much clearer separation of matches and non-matches! Such figures imply a reduction in operational cost of up to a factor of 46 depending on data quantity.
In the coming months, idetect’s data scientists—led by Dr. Holger J. Pletsch, Head of R&D—are going to carry on benchmarking these numbers with other competitor tools and publish results accordingly.
“idetect technology is the vehicle to achieve both compliance and operational risk mitigation quickly and more cost effectively than ever before possible,” said Olivier Merlan, Deputy CEO of Logos ITS. “It was high time to solve the problems that financial institutions have been facing for years. The immaturity of the technologies available at the beginning of the 2000-2010s did not make it possible to obtain results that met the high expectations of reducing the costs of managing transactional filtering and due diligence. Today it is inevitable that this sector turns to the technologies of tomorrow rather than remaining in a stagnant state using algorithmic methods of the early or mid-20th century!”
Mr. Merlan continued, “The obsolescence of technologies used for compliance purposes is an established fact, the importance of which I am in no way exaggerating: the vast majority of solutions used to-date still rely on outdated algorithms that generate unacceptable operational costs and compliance risks.
We wish to objectify the possible results in this area by quantifying the results obtained and by demystifying certain aspects of what years of misleading marketing have achieved. We already have scheduled proof-of-concepts with institutions that wish to objectively compare what we can bring with their historical solution. We are also in contact with leading market research companies and, without revealing our secrets, we will make available very concrete elements concerning our solutions.
It is our intention to allow financial institutions, and all players having an interest in name matching, to compare idetect with the historical solutions used for years and especially those that are claiming to be leaders in this area and to establish the facts. Sincerely, let the best win.”