Mujahid Merchant

Professional Summary

AI-Driven Climate Financial Risk Researcher | Springer & IEEE Author | Judge, Peer Reviewer & PC Member
5 years at Emirates Islamic Bank (Dubai) in credit operations and operational risk, recognized as a 2x GEM Award winner for operational excellence.
Published author with Springer chapters (2025, 2026) and active writer on Medium covering climate risk and mortgage credit markets.
Peer reviewer and program committee contributor across IEEE and international venues.

I build machine learning systems that quantify how climate disasters affect mortgage credit risk — and who bears the cost. My research uses causal inference, explainable AI, and gradient-boosting methods on real-world U.S. loan portfolios and federal disaster data to produce findings that regulators and lenders can act on.

My field is AI-Driven Climate Financial Risk: applying state-of-the-art ML to the intersection of climate hazards and credit markets. This is a small, fast-growing area where central banks (the Fed, ECB, Bank of England), GSEs, and Basel regulators are actively seeking methods to stress-test mortgage portfolios against physical climate risk — but where rigorous ML-based research remains scarce. I work in the narrow intersection where climate science meets machine learning meets mortgage finance, using real data from Fannie Mae, Freddie Mac, and FEMA to bridge that divide.

My methods include Double Machine Learning, Causal Forests, CatBoost with SHAP/LIME explainability, and DiCE counterfactual analysis. My datasets span 622,000+ loan-period observations across 2,827 U.S. counties, 200,000 Fannie Mae originations, FEMA disaster declarations, National Risk Index flood scores, and NFIP claims data.

Alongside research, I contribute as a peer reviewer and program committee member for IEEE and international conferences, supporting rigorous evaluation of manuscripts across AI, ML, and climate finance. I bring a strong execution mindset from banking, with hands-on expertise in credit risk evaluation, default indicators, and collateral assessment — domain knowledge that directly informs my ML research on mortgage default prediction.