Our Customers in Risk Management like Fraud Managers, Credit Managers and Risk Managers are all facing the same challenge to reduce risk of fraud and credit loss. Questions that they are asking Crystalloids are:

“How can I detect fraudulent transactions and claims?” ;

“How can I avoid accepting high-risk credit applications?” ;

“How can I avoid being left with unsold stock?”

The answer to these questions is embedded in our Risk Intelligence Solution.

Risk Intelligence is the discipline within Risk Management that uses Advanced Analytics to reduce risks in core, day-to-day, business processes. Each transaction, claim or application is an opportunity to detect fraud and to reduce risk of credit loss.

Risk Intelligence is the key to provide the necessary knowledge to reduce these risks. It is the process of collecting risk related data, analyzing risk patterns, exploiting risk models and evaluating these models.

Collecting risk related data

Vital to Risk Intelligence is risk related data. The more transaction, claim or application data you collect the more knowledge you gain on fraud and credit risk. But collecting data is not enough, it needs to be transformed in what we call a 360 degrees enterprise view. The enterprise view contains transaction, claim and/or application characteristics and customer characteristics to provide a total picture of the customer and related transactions, claims and/or applications.

Examples of customer characteristics are age and gender and more advanced characteristics are credit rating and average transaction amount per month. The enterprise view can range from 100 to 1000 or more characteristics, depending on the complexity of the organization. Having an enterprise view enables you to analyze the behavior of your customers and their transactions, claims or applications.

Technologies that we use to collect transaction, claim or application data are databases, data warehouses, ODS (Operational Data Storages), ETL (Extraction, Transformation and Load) and SOA (Service Oriented Architectures).

Analyzing risk patterns

Past behavior is the best prediction for future behavior. If you want to estimate the credit risk or accept an application for a credit card you best can look at your customers that already have a credit card and their paying behavior. By using advanced analytical applications on this historical data creditworthiness of applicants is analyzed to come up with profiles of high-risk applicants and low-risk applicants including the level of risk.

Technologies that we use to analyze customer behavior are advanced analytical applications like Statistical Analysis, Forecasting and Predictive Analytics with techniques like association, classification and segmentation.

Exploiting risk models

Risk models alone are not enough, it must be made actionable together with relevant business rules. These models and rules need to be scored in every transaction, claim or application. The scoring can be done in real-time or in batch. An insurance company for example can score incoming claims on the probability of fraud by using a fraud model in real-time. Claims with a high fraud score are automatically sent to a special investigation unit. Claims with a low fraud score can be handled automatically. A credit card company for example can score incoming credit card applications to determine the creditworthiness of each applicant. A low credit score indicates a high risk and will lead to an application being rejected. A high credit score indicates a low risk and will lead to an application being accepted.

Technologies that we use to score models are decision management in real-time or in batch.

Evaluating risk models

An organization that uses risk intelligence needs to measure the results of each risk model. You need to learn from every scoring of your risk models and business rules. Only then will you become a Predictive Enterprise.

Technologies that we use to evaluate customer response are Standard Reporting and Querying applications like Reporting, Querying, Slicing and Dicing, Dashboards and Alerts.

Read more on Risk Intelligence in Our References.