Robo Underwriting Model
For a Large Bank
CLIENT & PROBLEM STATEMENT
- A Large bank in the Private sector with a large customer base.
- A relatively new credit card business but very fast growing.
- The Client wanted to build a Credit scoring model for Credit cards for card granting decisioning on the go.
- It required automate the process of credit card underwriting by using applicants PII data & underwriters’ comments data of past applications.
APPROACH
- G-Square analyzed the demographics, various scores of applicants & textual comments & identified most important factors which will help in the Rejection/Approval of applications.
- We looked into various classification ML models plus text mining models to arrive at the final model for approval/reject process.
- Using our text analytics libraries, G-Square developed a robust Credit Underwriting Analytics Model using Machine Learning.
SOLUTION & OUTPUT
- Final outcome were APIs delivered for the decision of Approved/Reject/Need more info for the on the spot decision.
- The automated underwriting process is scalable, allowing the bank to handle an increasing volume of applications without compromising efficiency or accuracy.
