Case Studies

Collection Model And Strategy

Developing multiple collection models and an analytics driven collection strategy for a leading NBFC.

Customer Risk Assessment Using Alternate Data (SMS Data)

Developed customer risk assessment model based on non-personal SMS data with the help of NLP on raw text SMS information. Model was developed after creating 5000+factors on this data.

Digital Lending – Within 3 Minutes Loan Disbursement

End to end digital lending including finalizing customer journey, developing multiple predictive models (acquisition models, fraud models, income estimation models, SMS based models, etc.), approval strategy (score cut-off, policy rules, risk based pricing, limit assessment, etc.).

Bureau Score Development

Leading credit bureau wanted to launch its new bureau score for helping financial institution in assessing risk of customers. Developed segmented bureau score models which has been currently used by 200+ Financial Institutes.

Income Estimation Model

Developed models for estimating income of customer using bureau information. The model was developed for leading credit bureau to help financial institutions in assessing income of customer without asking for income proof.

Corporate & SME Rating Models

Developed 15+ industry specific corporate rating models (Manufacturing, Trading, Services, Contracting, Real Estate, Bank & FI, HNWI, Project Finance, Object Finance, etc.) for $200+billion asset size bank.

IFRS 9 Model Development

Developed PD, LGD, EAD, Macro-economic models along with impairment policy for IFRS 9 compliance

Portfolio Risk Assessment

Helped one of the top 10 global PE firm in risk assessment of a NBFC for M&A deal. As a result of risk assessment, PE bought significant equity stake in the firm.

Campaign Management

Optimized marketing email campaigns by identifying most appropriate keywords for email subject and timing of the day for maximizing email open rate.

Portfolio Valuation for Sell-off/Buy-off

Helped leading financial institution in identifying correct price at which they should sell NPA portfolio to ARCs. Estimated expected recovery of the NPA portfolio to access the price.

Acquisition Fraud Scorecard

Helped bank in identifying fraud applications by developing acquisition fraud scorecard which helped bank is reducing fraud loss as well as operational cost.

Next Best Offer

Developed next best offer models for a bank for identifying customer’s needs for providing relevant product targeting. Next best offer models helped bank in saving significant marketing cost as well as improved customer experience as customers started receiving relevant offers only.

Supply Chain Finance Models

Developed industry level Anchor, Supplier, Distributor and Retailer models for supply chain financing.

Sentiment Analysis

Helped fintech in identifying negative feedbacks of customers out of more 200,000 feedbacks for accessing areas where improvement or changes needs to be done.