Paris
The flexible decision engine for risk management
PARIS
The decision engine system
The PARIS system is a flexible decision engine suited to application scoring, strategy implementation and behavioral tasks
PARIS provides an easy to use point-and-click framework to implement scorecards, decision trees and assessment calculations. Intuitive tools allow you to create complex strategies and simple-to-implement scorecards allowing customers to be assessed for risk, attrition, profitability and other criteria. External information can be gathered at any point in the process and data on assessed applications can be stored internally or transmitted to any available database.
With state of the art XML data interface coding, PARIS is also .NET ready and easily configurable to be part of your online application system.
-
Scaleable processing
-
Central repository of:
-
Statistical models
-
Strategy trees
-
Customer data
-
-
Supports:
-
Multiple products/decisions
-
Champion/Challenger
-
Complex calculations
-
-
Multiple Strategy paths
-
Flexible front end to change scorecards
-
Integrated management information suite
-
Easily modified to suit changing requirements
-
Integrated development & production environment
-
What if calculations
-
Self-auditing
-
Unrivaled for flexibility and support
The diagram below illustrates a standard configuration of Paris. Data delivery mechanisms send data decision requests and associated data over a variety of possible delivery channels. The server-side mechanisms store account/customer information, make decisions by applying scorecards and rule-sets and send information (e.g. credit limit) back to the supplying system.
Multi-Route (Segmentation Process)
Marketing strategies, champion / challenger, acquisition channels etc.
Multi-Outcome
Links to other
Systems
Credit risk, Fraud risk, limits, authorisations etc.
API for bespoke links
User Configurable
Decision Trees/Rule Sets
Champion/Challenger
Flexible system allows on-site strategy and system design.
To control decision flow and apply rules.
Random or stratified sampling.
