Role: Senior Product Manager, SMB Lending — Khatabook (10M+ MAU merchant app), Aug 2023–Mar 2025.

The problem: Grow monthly loan disbursals on a 10M+ MAU merchant app without loosening risk or breaking under the weight of new lending partners. The three obvious levers fight each other — push growth and NPA rises, tighten risk and conversion drops, add partners and the system turns brittle. All three had to move at once.

The framework — attack growth, risk, and infrastructure as one system:

1. Funnel: Analysed where borrowers dropped in the credit-eligibility journey, paired drop-off data with borrower and ops feedback, and rebuilt the eligibility funnel, offer presentation, and onboarding flow.

2. Underwriting (negative-filtering): Instead of blindly relaxing rules, I studied how each underwriting parameter actually affected NPA, found several that filtered out creditworthy borrowers without protecting the book, hypothesised relaxing them, and validated through controlled experiments — each a ~500-loan cohort tracked over 3 months — rolling into policy only where delinquency stayed in range.

3. Infrastructure: Architected a partner-agnostic lending platform, owning both the LOS (partner API integrations, KYC, a configurable underwriting rule engine, autopay/e-mandate) and LMS (collections, reconciliation across user, payment gateway, and banking partner) — with validation checks and full audit logs, so a new partner's underwriting policy onboards as configuration, not code.

Results:

• +71% monthly disbursals (₹70 Cr → ₹120 Cr) in 14 months

• - 90% of loans disbursed within 3 days of KYC completion

• - -4% borrower drop-off and +16% average ticket size from the funnel rebuild

• - 4 NBFC/bank partners onboarded onto reusable infrastructure

• - 60% adoption of a co-launched embedded credit-protection product (new revenue line)

• - Led a 10-person cross-functional pod plus Risk, Legal, Compliance, and Partnerships stakeholders