This financial services company provides background information on retail and institutional account holders to more than 4,000 financial institutions worldwide. Specifically, this company alerts its clients when new or existing account holders are found guilty of finance-related criminal activity or are included on lists published by the Office of Foreign Assets Control (OFAC) of the U.S. Department of the Treasury (and similar lists published by foreign governments), which prohibit dealings with individuals and companies owned or controlled by, or acting for or on behalf of, targeted countries, and other individuals, groups, and entities, such as terrorists and narcotics traffickers.
Most of the company’s internal business processes for identifying and recording this data and for alerting its clients were manual. The Vice President of Operations wanted to dramatically shorten the processing cycle times and improve the quality of its data outputs by automating its core processes with new technology.
Iknow was chosen to design, develop, and deploy the company’s advanced technology platform. Iknow was selected because of its systems integration experience and knowledge of the leading text analysis products.
The first phase of this effort focused on the data collection, data extraction, and data quality tasks. The approach involved four important steps.
- Target source identification and clean-up. Iknow worked with the company to update their list of source content, including removing inactive content sources and updating their content licenses.
- Build the connectors. Iknow built the individual connectors to enable the platform to collect information directly from more than 600 federal, state, and local government websites and indirectly from thousands of news articles published every day.
- Create customized rules. Iknow built more than 200 custom rules for extracting names and other entities and for validating the extracted entities. This customization is essential because certain name spellings, complex sentence structure, punctuation variations, and other language-related problems cause the software to miss or misinterpret the desired content.
- Load the outputs into the existing database. The extracted entities are converted into structured data and loaded into the company’s existing proprietary database.
Iknow designed and implemented the complete system. The SAP BusinessObjects Text Analysis software was chosen as the core processing engine for natural language processing and entity extraction functionality. The system also captures and stores the full-text articles to provide back-up support for every entry in the database. In addition, Iknow performed overall system quality assurance testing and validation, prepared and conducted end user and administrator training, and prepared the technical documentation.
The system’s output is captured in the company’s proprietary database of names of individuals and companies who have been convicted of finance-related criminal activity by regulators and the exchanges. The cycle time for data collection, analysis, processing, and loading is as short as five minutes. The Vice President of Operations summarized the new system by stating “The new platform enables the company to conduct comparisons of clients’ new account information with near real-time data in our proprietary database. We report back to our clients on the same day with any information on those accounts that requires further examination."