Back to work

Product · Case studies

Case studies

Deeper strategy and teardown work, each framed by the problem, what I did, and what changed.

Payments Ops · Build

PayOps Copilot: payment reconciliation

A reconciliation workspace for payment-ops teams: rule-based settlement numbers, AI limited to investigation drafts, and a full audit trail.

7
exception types
3+
gateway formats
AI-bounded
rule-based core

Problem

Payment-ops teams reconcile across three inconsistent sources: internal orders, gateway exports and bank settlement reports. When numbers don't match, it's unclear whether an exception is real or just a schema mismatch, who owns it, and when it's due. The workflow is spreadsheet-based, with no accountability.

Approach

  • Ingest and normalize multi-gateway CSVs (Razorpay, Cashfree, PayU, generic) without silently dropping rows.
  • Match on order IDs and gateway references, compute net settlement after fees and GST, and flag seven exception types: missing or duplicate captures, missing settlements, amount mismatches, refunds and chargebacks.
  • Keep the math rule-based; AI only drafts the investigation notes, and never calculates the settlement numbers or moves money.
  • Turn exceptions into org-scoped cases with RBAC, SLA timers and a full audit ledger.
ReconciliationSettlementStructured OutputRBAC · SLA · Audit

Product Strategy

Airbnb India: Flexi-Booking

A strategy to solve host revenue instability and guest booking inflexibility, built around tiered insurance and a rewards / credit system.

−35%
cancellation losses (potential)
+15%
host satisfaction
+10%
booking conversion

Problem

Hosts faced unstable, unpredictable revenue while guests wanted more flexible cancellation options. That tension eroded trust and hurt conversion on both sides of the marketplace.

Approach

  • Designed a tiered insurance layer (AirCover) plus a rewards program to reduce host revenue risk.
  • Introduced flexible-cancellation incentives via a credit system to build host–guest trust.
  • Prioritized the roadmap with RICE to focus on the biggest opportunities.
Product StrategyRICEInsurance / AirCoverTrust & Safety

Product Teardown · Product Space

Zepto: AOV & Retention Teardown

A teardown analyzing how to lift average order value and retention through better product discovery and engagement.

₹460→₹600
target AOV
45%
retention focus
Top 5%
awarded at competition

Problem

Zepto needed to raise order values and improve a ~45% retention rate without compromising its fast-delivery promise.

Approach

  • Analyzed purchase behavior to find levers to move AOV from ₹460 toward ₹600.
  • Proposed enhanced product-discovery and engagement mechanics to deepen retention.
Product DiscoveryRetentionEngagementTeardown

Want the full walkthrough?

Get in touch