SAS: An Essential Layer of Security Intelligence Protection

SAS: An Essential Layer of Security Intelligence Protection

CIO VendorJim Goodnight, Co-Founder & CEO
Every fraud begins with a belief—a belief in breaking the security in place. The definition of fraud has changed with time, and advancements in technology and cyber-enabled fraud are what driving the modern-day fraudulence. Globally, organizations lose approximately 5 percent of their annual revenues to fraud, which translates into potential losses in excess of $2.9 trillion worldwide. While direct losses due to cyber fraud are staggering in dollar amounts, the actual cost is much higher in terms of loss of customer confidence and productivity. Jim Goodnight, the godfather of data analytics and the co-founder and CEO of SAS explains how most of the existing fraud detection systems typically provide reactive, after-the-fact analysis of questionable transactions—an approach that is too slow to give any real protection from loss, making real-time transaction scoring their best option for preventing a fraud.

Based on this discovery, SAS®—one of the world’s leading business analytics software vendors—offers an enterprise-wide strategy that unifies fraud and security systems, focused on improving detection and prevention efforts. By harnessing the collective wisdom of fraud managers across the world, SAS designed a solution that captures the practical principles of fraud management and can be specifically tailored for each local implementation. The company provides its fraud detection and prevention capabilities to various financial institutions such as banks, capital markets, and insurance companies.

Detecting and Preventing Fraud across the Enterprise

SAS® Fraud Management—a full-service fraud management solution—provides a holistic, customer-centric view of fraud that is unrestricted by transaction type or business channel with the capabilities to monitor multiple lines of business from a single platform. SAS Fraud Management offers 100 percent real-time scoring and decision capabilities by looking at payments, fund transfers, purchases, and non-monetary transactions, making it a system that provides highest breadth of coverage. Also, SAS Fraud Management provides holistic and real-time integration of advanced analytics, alert management and reporting, authorization systems, rule writing, simulation capabilities, and on-demand scoring. “Our advanced analytic technologies significantly reduces the number of false positives and increases fraud detection rates,” states Goodnight.
The company considers fraud prevention as a dynamic process with no start and end point. SAS’s risk/fraud detection models are built by combining individual and collective bank data, which is then used for scoring and decision making. Having considered end-to-end fraud prevention, the solution also includes a highly functional fraud alert management and investigation system, a flexible connector to connect with authorization systems, and extensive reporting capabilities. Additionally, SAS’s alert management capabilities helps clients to make more effective use of staff resources and reduce time and money investigating false positive by uncovering greater operational efficiencies. SAS Fraud Management can be used as the primary fraud detection and investigation solution owing to its real-time operational and performance reporting capabilities.

Combating Enterprise Payments Fraud

SAS Fraud Management helps clients to connect the SAS solution with their processing and authorization systems using a universal SAS connector (USC) interface kit. The transactions are then sent to the USC from the authorization system, where they are appended to the appropriate data based on the transaction type. This data contains information such as user and system variables and varies depending on the transaction type. The union of SAS scoring engine and the USC allows users to control the firing of models and the execution of the user-written and system rules.


SAS Fraud Management offers 100 percent real-time scoring and decision capabilities by looking at payments, fund transfers, purchases, and non-monetary transactions, making it a system that provides highest breadth of coverage

This is then followed by the scoring engine, which executes appropriate fraud models and the associated decision logic specified by the user to produce a score.

Profiling 100 percent of all transactions enables companies to capture key customer behavior data. The solution also allows users to develop and create customized profiles using public signature entities such as account, customer, beneficiary, and IP that are under the complete control of the client organization. In addition, SAS Fraud Management includes a set of SAS-built profiling and scoring models to assess and monitor the payment risk associated with accounts and customers. The solution also leverages the industry’s most advanced analytics to give banks a significant advantage in the quest to manage and control their fraud losses. Along with its improved proprietary fraud modeling techniques, SAS’ innovative hybrid modeling approach provides the benefits custom and consortium approaches. The solution incorporates a banking client’s data as the core model, and then integrates pooled data to create a more predictive single model with the power of a consortium model, while retaining the value and flavor of the company’s customer data. While mostly algorithmic methods are used to express customer behavior, SAS uses the power of raw data to monitor every single transaction. Helping financial companies understand customers’ overall relationships with their bank, the solution’s unique “signatures” approach captures customer behavior patterns from every source and evaluates that information every time a transaction is requested or completed.

A Hybrid Approach to Surveillance

SAS adds a proactive layer of security intelligence protection by leveraging a unique blend of advanced analytics, embedded artificial intelligence, and machine learning, safeguarding every client's reputation and bottom line. Providing a holistic view of an organization's fraud risk, AI techniques, such as unsupervised intelligent agents, identify emerging threats and automatically suggest new rules and scenarios in real time. Focused on effectively capturing information on data irregularities, SAS’ enhanced model performance is credited to their proprietary modeling technique—Self-Organizing Neural Network Arboretum (SONNA). By utilizing an adaptive segmentation scheme, SONNA goes through a continuous evolution during the model building process based on its ability to detect changing customer behavior. This not only leads to significantly increased fraud detection performance, but also enables the models to become more robust over time. In addition, SAS Fraud Management also provides superior capabilities such as rule management configuration, alert management, and a full reporting system that allows users to easily generate reports using any data in the system and publish them in a variety of formats, including HTML.

What makes the company second to none is its hybrid approach to surveillance that covers social network analysis, graph analytics, anomaly detection and text analytics, powered by a mix of AI techniques, advanced analytics, and machine learning along with traditional detection methods to discern and mitigate suspicious events across all lines of business.

Carving a Secure Future

SAS is currently leveraging its domain expertise, analytics and technology to reduce losses from modernized compliance, identifying threats to public safety and national security, fraud, and improper payments. The company is working on building its next phase of the integrated platform with a foundational architecture that can support modular expansion—for both AML and fraud—on a single offering. They also plan to design and offer fraud detection engines suited specifically for meeting the business needs of every client.