A large commercial bureau wanted to build a tool which would help commercial banks and agencies scrub through the application and understand its propensity to fraud.
We helped in the design and development of the fraud detection tool by combining various data sources like commercial bureau information, creditor information and past search data.
The overall solution comprised of a phased approach which involved:
- Data Collection
- Data Preparation
- Matching and Patten Analysis
- Model Design/Build
The outcome was system consisting of Fraud Repository and Inquiry Database holding historical information from multiple sources based on past searches, related demographics, and available bureau information.
The overall Information flow happened across three phases - Application Process, Match and Report Generation. The Match process performed a search on the Database to identify applicant’s profile from past data. Finally, the Report was generated on applicant’s fraud risk credentials.