S

SmartMatch

Solving Uncertainty in Lending

Carousel Gallery Item: 1
Carousel Gallery Item: 2
Carousel Gallery Item: 3
Carousel Gallery Item: 4
Carousel Gallery Item: 5

Last updated: 01 August 2020 04:29 AM

Created project

The problem SmartMatch solves

Problem Statement
Millions of customer interactions have helped us identify the most critical concern of borrowers : “From where will I get the Loan?”. On the other hand, lenders assess the applicant’s intention and ability to repay. The absence of an optimal match-making solution leads to a painful experience for the borrower and high costs for lender & platform.
Objective
Develop AI-ML based SmartMatch, matching borrower to the lender based on his income and profile, borrowing and repayment history
Context
Typically, customers end up applying to a lender basis brand and lowest RoI. In a lenders’ market like ours, their criteria always remains a secret, and hence, 50%+ apps get rejected.
With 7000+Cr of disbursal over 6 years, SmartMatch is a self-adjusting advanced predictive algorithm that matches a borrower’s profile with the lenders’ criteria, helping customers apply to right lender in first time.
Challenges
40-60% Approval rates with 80% rejections due to unfavourable credit profile or financials
Every rejection technically reduces the chances of user getting a loan from another lender also, because of reduced credit score and increased enquiries. Hence, more applications keep reducing the customers’ chances, leading to eventual ineligibility.
Decisioning needs bureau and banking data. While the former is standardized and structured, banking data is not easily accessible and in an un-standardized format, hence not scalable.
Implementation
User enquires about Personal Loan by entering basic details
Fetch Credit bureau data & access AA to verify customers’ banking details
SmartMatch ingests all data to provide a real-time score for all offers & rank order the offers
Customers are advised to select the highest ranked offer
Final app + Docs are shared with the lender for decisioning
Expectation
-Increase in Approval rates by 25%+ due to ‘right offers’
-Reduction in TAT by 20%+ due to ‘FTR’ apps
-High efficiency -> significant cost reduction for lender & partners

Challenges we ran into

Seamless New User Registrations: New users are unable to create a handle on an FIU platform in the absence of an AA SDK. They have to be redirected to an AA app to register.

Technologies used

Discussion