Krishnendu Saha Email & Phone Number

Gurugram, Haryana, India

Gurgaon, India

Strategy - Marketing & Acquisition Analytics• Developed and implemented the following strategic acquisition controls to weed out:• Customers with offer misuse history with AMEX, potentially savingaround $20M+ in annual acquisition costs• Early Tenure applications, potentially saving around $30M+ in annual acquisition costs• Cross-sell/Same card applications with historically poor KPIs, potentially saving $60M+ in annual acquisition costs• Re-calibrated deterministic controls to exempt highly engaged AMEXcustomers from probabilistic models, potentially saving around $10M+in annual acquisition costs• Responsible for baking in latest enterprise level Acquisition Economicsto ROI linked controls• Led several initiatives to re-direct Acquisition costs from unprofitableto more profitable acquisitions during uncertain macro-economic scenarios• Identified a critical Offer Gating Issue in one of the Co-Brand DirectMail campaigns, which had led to serial offer abusers getting acquiredGovernance - Credit Bureau Analytics• Responsible for the end-to-end outbound reporting analytics of USConsumer and Small Business to Credit Bureaus• Developed new Credit Bureau Control Oversight MIS to efficientlytrack down reporting issues• Drove significant customer impact (10M+ US customer accounts)across several key projects (Scrubbed SSN Project, US Stale Trades remediation, Credit Bureau Rejects etc), by making the bureau reportingprocess more accurate and efficient• Developed 10+ KRIs to identify inaccurate bureau reportingModeling - Credit Risk Analytics• Responsible for developing and upgrading the proprietary Credit Scoring Model of American Express• Developed critical Credit model features during COVID by leveragingexternal credit bureau triggers to provide moratorium benefits to customers• Identified $20M+ non-profitable risk pocket in US Consumer portfoliothrough case reviews of high balance default cases

Gurgaon, India

Developed a Fraud Detection Engine for detecting fraudulent applications in the online credit lending process, using various fuzzy string matching and manually engineered algorithms. Removed field investigation for 40% of the disbursals and improved the overall credit lending efficiency of the firm by 30%.