It is increasingly compulsory for enterprises to be vigorous in efforts to ensure their customers are not a source, destination, or conduit for illegal funds, such as those derived from crime or terrorism. Today millions of alerts are generated by technologies that only address one part of the problem. We propose something radically different from anything before it.
Eliminate the risk of involuntarily facilitating crime, avoid fines and Reputational Exposure.
Reduce considerably the amount of labor associated with watchlists and alerts handling.
Use technological advances to refuse status quo and make a major step forward.
Consume data from a wide variety of sources without volume, variety and velocity constraints.
Manage the complex interaction between providing dynamic compliance detection and avoiding generating huge alert volumes that are operationally expensive to manage, resulting in unproductive resource consumption.
Contrary to regular technologies, iDETECT® analyzes pieces of corroborating information, or alternatively one highly discriminatory piece of information rather than a threshold-based model. Our model ensures that desired matches can be found without driving a corollary spike in false positives.
There is no such thing as a single detection rule which detects all persons and entities that an enterprise would want, or a regulator would require. Therefore, iDETECT® was built to ensure there are no limits in the number of attributes that can be analyzed to resolve identities.
Benefit from a flexible and inexpensive mean to capture watchlist information, directly from the internet or your proprietary sources, through a simple automated crawling and importing process.
The automated iDETECT® crawler browses the World Wide Web methodically to find information in near real-time. Through a regular check of Websites or Rss feeds (e.g. Interpol, EU Sanctions, Law Enforcement lists, Newspapers, Blogs, …) the crawler obtains the most comprehensive, recent, and accurate data possible.
iDETECT® identifies the entities and their context through "text mining". Return-on-investment is immediate as our approach does not require the use of payable watchlists.
A typical watchlist can contain many hundreds of thousands of records. More records means more probable matches, and more false positives. Clearly, from an investigation perspective, the attention shall be on the higher risks, but all risks, even regular-level risks, must be verified.
Our approach looks at customers and their relationships in a variety of characteristics and circumstances, and allocates them on a risk basis. As not all of them represent equal risk, it is appropriate to identify those which represent the highest risk, deal with these first, and so on. By leveraging contextual identity-based analytics, machine-learning, and peer group modeling watchlist monitoring become easier and more accurate.
Different matching algorithms are suitable for matching different forms of data, and there is no algorithm that is fit for all forms of data (e.g. structured data, free text, numeric data, date information,…).
iDETECT® provides the largest available set of matching algorithms and transliteration capabilities to manage multicultural name data sets.
Whilst leveraging linguistic rules that are associated with named-entities, our technology can match structured data with structured data, structured data with unstructured data ... and even unstructured data with unstructured data! Using Hadoop®-based Distributed File System for big data volumes, there is simply no software in the watchlist monitoring industry that has provided such capability before.