Muthukkumaran K

Inside a large retail lender, a collections challenge initially framed as a targeting problem revealed a wider operating constraint: customer selection, payment friction, campaign overlap, and field escalation were being managed separately.

Muthukkumaran designed a low-cost intervention across an approximately one-million-customer retail portfolio, combining customer-level selection, BBPS-enabled payment journeys, clearer communication, suppression windows, and controlled escalation. Lightweight scripts and existing teams tested the economics before technology scale-up.

The intervention doubled digital collections within six months and delivered INR 2.2 crore in net savings over one full year. Across the wider fintech initiative portfolio he led, equivalent net savings reached approximately 4% of the lender's full-year PBT.

Inside a major private bank, Muthukkumaran drove a new-to-market online seller-finance platform across product, credit, policy, legal, information security, technology, operations, ecommerce partners, and a multi-vendor delivery chain.

He designed the end-to-end process, secured process approval on first submission, and led a 15-member cross-functional war-room through two UAT cycles covering more than 3,000 cases. Approximately 70% of the processing stages were provisioned online.

The cloud-data proposal also cleared information-security review on first submission. The build created reusable patterns for later digital-lending products: cross-functional governance, vendor orchestration, and exception-aware workflow design.

In a consumer-finance business processing more than 100,000 applications a month, Muthukkumaran led an AI underwriting program across the product team, internal analytics, technology, the loan-origination-system provider, and a specialist AI partner.

Historical model building and back-testing produced a decision frontier from which the business could select an operating point. The validated point targeted a 10-percentage-point increase in approvals with a 3-percentage-point reduction in bad rate.

That validation became a key input to a USD 7 million investment in the partner, after which he led three-party integration of the model into the underwriting flow for an on-field pilot. The experience reinforced a discipline he carries into IDAI: model potential, implementation progress, and realized production value are different claims, and each must earn its own evidence.