Decisioning Platform for Real-Time Multi-Engine Fraud Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing financial transaction card networks face challenges in integrating multiple vendor fraud scoring products effectively, lacking flexibility and configurability to provide value-added services to customers, and require a system that can easily combine and orchestrate these products for enhanced fraud risk assessment.
Innovation Solution
A computer-implemented method and system that enables the integration of multiple fraud scoring engines, allowing for real-time fraud prediction and customizable fraud scoring models, with a plug-and-play architecture that supports various input and output channels, and enables real-time fraud management and case management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple vendor fraud scoring products are integrated, then fraud prediction capability is improved, but system complexity increases
Solution Approach 1:
The system segments the fraud scoring functionality by implementing separate scoring engines for different fraud types (account takeover, synthetic identity, credential stuffing, etc.). Each scoring engine operates independently and can be configured separately, allowing the system to handle multiple fraud prediction tasks without overwhelming complexity in a single monolithic system.
Solution Approach 2:
The decisioning platform is designed as a universal system that can accommodate multiple scoring engines and evaluation methods within a single framework. The platform provides common infrastructure for data collection, model execution, and result aggregation, allowing it to serve multiple fraud prediction functions while maintaining manageable complexity through shared resources.
2Adaptability or versatility
If multiple scoring engines are combined, then value-added services are improved, but ease of operation deteriorates
Solution Approach 1:
The system implements dynamic configuration capabilities where scoring engines can be activated, deactivated, or adjusted based on specific transaction types, customer profiles, or risk thresholds. This dynamic nature allows the system to adapt to changing operational requirements without requiring complete system reconfiguration, thereby maintaining ease of operation while providing versatile value-added services.
Solution Approach 2:
The decisioning platform acts as an intermediary layer between various scoring engines and the final fraud assessment decision. This intermediary component consolidates outputs from multiple scoring engines and presents unified recommendations, simplifying the operational interface for users while maintaining the complexity of multi-engine processing internally.
3Productivity
If real-time processing is implemented, then productivity is improved, but device complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-loading scoring models, risk thresholds, and evaluation criteria into memory before real-time processing begins. Historical data and model parameters are prepared in advance, allowing the real-time decisioning process to focus only on executing the scoring engines against current transaction data, thereby achieving real-time processing without the complexity of real-time model training or data preparation.
Data Source
AI summary
A computer-implemented method of providing enriched transaction data for a transaction requiring an authorization is provided, the transaction performed using a computer system having a processor and a memory device. The method includes storing transaction data received from an input channel, the transaction data including a transaction identifier. An execution plan is retrieved based at least in part on the transaction identifier. The transaction data is processed across an enrichment processor based on the execution plan to generate at least one fraud score for the transaction. The transaction data is enriched to include at least one of the fraud score and an enriched data object. The enriched data is transmitted to an authorizing party for authorization.