Services
AI Model
Artificial Intelligence (AI) Model and Rules Management Scoring Framework
Every transaction is evaluated by two systems:
AI Model
Model checks against multiple transaction indices and compares them to patterns identified in historical data that is correlated with fraudulent use.
Versus the Rules
The system checks against rules set by the customer within the Rules Management tool.
Types of rules:
- Score Rules: Used in the final score of the transaction.
- Action Rules: Direct outcome either accept or reject.
Scoring Logic of AI Model and Rules
Each System Generates:
- Their own score
- An accept/reject recommendation
- Reason codes and their details
Final Score and Recommendation
The final score and recommendation is determined by the Transaction Risk Management processing logic.