ABN Amro Fortifies Compliance Searches

By Maria Trombly | Securities Industry News | April 3, 2006

Dutch banking giant ABN Amro has put in place a compliance system with a high-tech search feature designed to improve information gathering for 400 compliance officers worldwide. The system's key feature is a Web-based portal that the bank calls Galileo, which scans unstructured content including regulations, legal files, internal documents, video, audio and e-mails for meaning and context.

ABN Amro based its system on the Intelligent Data Operating Layer server and real-time content analysis technology supplied by Aungate, a division of U.K. technology company Autonomy Corp. The server provides real-time monitoring of corporate communications and files, and it can be used to classify internal documents so that employees don't have to spend time entering that information.

ABN Amro needed a system that could recognize conceptual links between pieces of information, make staff members automatically aware of related compliance issues and deliver information while they were in the process of working on a particular case or business project, according to David Kemp, development and communications director at ABN Amro. He recently described the system at the Securities Industry Association's compliance and legal division seminar in Hollywood, Fla.

"For lawyers or compliance officers, the real benefit of the Aungate solution is that it can provide focused information on specific subjects they are interested in," Kemp said.

Aungate's intelligent search software scans and analyzes the content of e-mails, voice recordings and instant messages. Several securities firms use it for monitoring, alerting, policy management, electronic discovery and compliance. The company's customers include the Securities and Exchange Commission, New York Stock Exchange, National Association of Securities Dealers and Department of Homeland Security.

"Our technology allows them to look for information that they don't know they need," said Ian Black, managing director of Aungate's Washington, D.C. office. "You can't search for things that you don't know about, and you can't pull up a conversation that doesn't use any of your key words."

Compliance today requires strong search tools capable of understanding content. The text in, say, an e-mail message can be related to outside events. For example, the NYSE compares e-mails to trading data and news flows and looks for correlations between them. "If someone is trading ahead of news, we can determine if they're an insider," Black said.

ABN Amro plans to use Aungate technology for tasks such as monitoring internal communications. Kemp said, "The system has been acquired because of its scalability. The possibilities here are quite extensive."

Another Aungate client is Lynbrook, N.Y-based Doar Litigation Consulting, which specializes in document discovery for legal proceedings. Doar formerly relied on conventional keyword searching, said CEO Samuel Solomon. The new platform embraces intelligent, context-sensitive searching that understands the meanings within search phrases and target documents.

Securities firms involved in class-action litigation cases have to deal with an extraordinary amount of data as part of the discovery process, said Solomon. "Traditional methods of search and analysis are becoming more and more difficult in terms of cost management and effectiveness," he noted. "You cannot use traditional search methodologies, especially in the early parts of potential litigation."

The Autonomy-Aungate technology is "agnostic to the words themselves," said Solomon. It looks for suspicious associations between phrases that don't normally belong together. This requires that the search tool actually infer the meaning of what is said. "We wanted a tool set that would provide content analysis," Solomon explained.

"The heritage of [Autonomy] is serving intelligence agencies," Black said. Ultimately, compliance searches will be able to dispose of the familiar Web browser search box altogether. "The dependency on having a person go and look for something is terribly inefficient," Black said. "Very rarely do investigators find things that they want to find. Instead, they stumble on things by accident. An intelligent search is going to know what you're looking for before you do--that's where most of our customers are going."

The next step, he said, will consist of intelligent applications that work together so that it matters less what form information is in. The meaning will be what matters.