Great Eastern Life Assurance (Malaysia) Berhad
Contact: Pang Ghee Jian
Country: Malaysia
Award Category: Innovative Problem Solver
Tell us about the business problem you were trying to solve.
Great Eastern Life Assurance Malaysia (from here onwards, referred to as GELM) is facing increasing challenges across the Life Claims Department of its business. GELM’s Request for Proposal for a Claims Fraud Detection Engine describes the challenges extremely clearly as follows:
At all times, Life Claims Department is encountering a range of fraudulent claims, including misrepresentation of material health info at policy application stage, exaggeration of health conditions, forgery of medical reports, identify fraud where someone’s personal identity information are deliberately used by syndicates to purchase insurance policies and gain financial advantage through false claims submission, impersonation at clinics and hospitals in order to obtain medical results in favor to the customers, and so on.
GELM is looking for increased efficiency in the identification of fraud and the development of better-quality alerts. The goal is to streamline the claims and application assessment process and take a proactive approach to detecting and deterring fraud in the business.
By collaborating with SAS, we took a unique, end-to-end approach to detecting, preventing and managing both opportunistic and organized fraud. The solution includes components for data management, automation of fraud detection, investigation and case management.
What SAS products did you use and how did you use them?
We use SAS visual investigator to detect and assess potential fraud cases. The journey begins with the BDP extraction, where a daily schedule is set for a complete base extraction. The BDP files are source like files, hence necessitating that the staging job applies unique logic for each source system. A combination of control and cleansing jobs will then extract data from the Vault schema, applying cleansing procedures to produce Cleansed Element Group (CEG) tables. These CEG tables store the cleansed data, which then form the input for the entity and network generation process.
In addition, there is an analytical view that includes links to networked data (entities), along with other remaining variables, to support the scoring framework. Then, in SAS Visual Investigator, the front end is tailored for users to view alerts and conduct investigations with data from tables like VI entity, VI sub-entity, and VI relation. When an alert is raised, users can swiftly access pertinent details, linking policies, claims, and individual entities for a comprehensive investigation.
What were the results or outcomes?
This solution detected cases and contributed RM 451k of cost savings in first 10 months of 2024.
Why is this approach innovative?
The solution is 100% automated and it has the features of network analysis which can connect claimants within the network for guided and detailed investigation.
