Sumit Kumar Gupta is a Quantitative Risk Model Development Lead with a decade of experience designing and deploying Default Risk and Loss Severity credit risk models—Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD)—for wholesale banking portfolios, which Underwriting and Risk Management depend on under U.S. regulatory frameworks including CCAR, CECL. His work sits at the intersection of traditional quantitative credit risk modeling, Federal Reserve SR 11-7 model risk governance, and the applied use of AI/ML in highly regulated environments.
Based in Tampa, Florida, Sumit has led model development and remediation across multiple CCAR/CECL submission cycles, engaged with Federal Reserve examiners and internal model validation teams, and authored regulatory model documentation across wholesale credit loss forecasting and commercial real estate models. He is actively engaged with the evolving model risk management landscape for Generative AI and agentic AI systems in financial services, including the implications of the recently issued SR 26-2 guidance.
Sumit holds an M.S. in Industrial Engineering with a major in Data Science from Northeastern University, Boston.