January 3, 2012

Regulations require financial institutions to establish compliance programs which must include: development of internal policies, procedures and controls; an ongoing employee training program; and an independent audit function to test those programs. Firms make efforts to comply with those requirements by deploying expensive data-intensive technologies, and they devote vast human resources to manage usually large volumes of alerts.

Compliance analysts use alert management tools and navigate through pull-down menus, choice boxes, radio-buttons, scroll-bars, and other interface elements of software systems to get answers to sometimes very simple questions. Very often those answers come delivered in pieces scattered over grids, charts, graphs that are typical components of modern software dashboards.

Can compliance questions be asked by people and answered by computers in a simple dialog? Could such an interaction look like an instant message or email dialog between two analysts working on the same case?

Texting, blogging, or SMSing have became the preferred style of communication among people. Twitter attracts millions of users, even if it has a strict length limit on text messages. People seem to be comfortable with this style of instant communication due to the natural language expressive power and simplicity. Can people communicate with computers in the same way they communicate with each other? The state of the current regulatory compliance technology indicates that this not yet the case.

Let's assume for a moment that a compliance analyst communicates with a software agent through texting or emailing inquiries or requests for investigative actions. The software agent then reasons about those input requests by using regulations, policies, and decision rules included in its ontologies and rule bases. Finally, the agent responds to a request with an instant message that includes analytical results and references to underlying regulations, policies, procedures, and data sources. The agent can take actions, if instructed to do so.

Software agents that provide specialized regulatory compliance services would need to operate in the service-oriented, semantically-enabled enterprise IT environments. Since they would contain repositories of regulatory knowledge and best practices, they could support or take over time consuming investigations of routine regulatory compliance alerts and cases.

Emerging semantic knowledge-based technologies offer intelligent reasoning capabilities. We should take advantage of those developments to represent semantically regulations, policies, procedures, investigative experience, and decision rules.

Using the semantic rule-based technologies Exprentis has developed the Regulatory Compliance Assistant and Regulatory Knowledge Base products. Read more how they can help you to make your compliance programs more efficient.