Credit Risk Scoring Engine

Objective

Build, Deploy and Productionize a combination of models for instantaneous loan qualification across Storefront and Online Channels

Business Problem

Risk Management unit of a Retail Lending business needed seamless solution for loan application underwriting of pay-day loan & short term (installments) consumer loan products for returning customers via omni channel touchpoints

Solution Overview

Customers can now apply installment loan/payday loan through web/physical branch
Credit Scoring Engine developed in collaboration with Risk unit generates PD (Probability of Default) score for internal teams and Denial response for customers, by processing loan parameters like Age, GMI, Previous Loan History, et.al
Further scores are run through set of business rules to decide the final loan amount for approved for applications

Output/Value Add

Improved efficiency and quicker turnaround in credit appraisal process
Risk based loan limit setting
Approval/Denial reasons are provided driving customer trust

Solution Architecture

Solution Brief

Our solution for credit scoring was using more data to provide an individualized credit score based on factors including current income, employment opportunity, recent credit history, and ability to earn in addition to older credit history. This Solution can also adapt to new problems, like credit card churners, who might have a high credit score. This Solution can also satisfy regulatory requirements to provide reason codes for credit decisions that explain the key factors in credit decisions.

Application Flow

Seamless integration across lifecycle of customers applying loan for both secured and unsecured products
Integrate the loan application for each customer to corresponding product and loan type
Enable Mashery to trigger API calls real-time
BW to process and queue the requests to Statistica
Statistica to score the loan application and return PD and Plausible denial reasons