Mar 2026 – Present · 2 mos

Sep 2023 – Mar 2026 · 2 yrs 6 mos

• Led the design and delivery of multiple AI‑powered risk and intelligence solutions at Mastercard that directly strengthened revenue, reduced fraud losses, and expanded customer value.
• Championed a Gen AI‑driven Risk Detection system by combining multiple data streams, significantly improving fraud capture while reducing false positives at a product level.
• Delivered an end‑to‑end merchant onboarding risk insights product that generated $23M in revenue, owning the journey from concept to production.

Sep 2021 – Sep 2023 · 2 yrs

• Merchant Risk Predict Product: Contributed to build product solution aimed to provide customers/partners with relevant and actionable probabilistic scores to help identify the most relevant fraudulent activity arising from a merchant, their compliance violations, their terminal activity, chargeback, and return risk they bring to the network
• Managed development and timely delivery of two AI-based models on predicting EMV Chip Failure and Anomalous Time of Transaction for over 500B merchants. Analysed KPIs & headed model improvement for real time scoring of transactions using exponential decay velocity features at DMP level
• Area-of-Use Circumvention: Executed end-to-end POC using stacked XG-Boost and Isolation forest anomaly detection models to detect customers operating outside licensed area of use
• Location Intelligence for Aggregate Merchants: Led team of 5+ members to develop intelligent solution to cater to aggregate merchants optimise the placement of their physical stores, drive-thrus and delivery centres in a city to maximise stakeholder specific business goals(improve branding, sales, minimise fraud risk etc.)

Apr 2020 – Sep 2021 · 1 yr 5 mos

• Anti-Money Laundering: Developed AI enabled solution using Deep Learning and Graph techniques to pinpoint customers indulging in suspicious money laundering activities. Solution saved >55 days manual work monthly for Compliance Team & reduced false positive rate of legacy rule-based system by 70%
• Digital Standards Product Recommendation: Created product for Issuer Banks to drive higher transaction approval rates for Card-Not-Present transactions. Used DeepFM architecture in Tensorflow for developing approval model and applied simulations to recommend Mastercard product(s) providing highest boost to customer’s approval rates. Scaled product pipeline to distributed system using Analytics Zoo, PySpark, BigDL for 3 more regions Europe, Middle East, and Canada

Oct 2019 – Apr 2020 · 6 mos