Director with extensive experience at KPMG, specializing in advanced credit risk model development and regulatory compliance. Proven track record in stakeholder engagement and model validation, enhancing accuracy and risk differentiation.
Adept at project management, driving impactful results through innovative methodologies and cross-functional collaboration.
Overview
11
11
years of professional experience
Work History
Gurugram
KPMG - Risk Advisory Services
07.2023 - Current
Development of Advanced Internal Ratings-Based (AIRB) credit risk models for diverse portfolios including General Industries, Project Finance, and US Pension Funds, ensuring alignment with Basel regulatory standards.
Remediation of ECB findings
Calibration and validation of Probability of Default Master Scale using firm-specific internal default data across multiple asset classes, enhancing model accuracy and portfolio risk differentiation.
Collaborated with cross-functional teams to address regulatory inquiries, providing detailed model documentation, impact analyses, and methodological justifications to support supervisory reviews and audits.
Remediated ECB (European Central Bank) findings by implementing targeted model enhancements, revising documentation, and aligning methodologies with regulatory expectations to close supervisory gaps.
Senior Manager
American Express
09.2022 - 06.2023
Designed and implemented Hyperparameter Optimization Policy Standards to ensure consistent, transparent, and auditable tuning practices across AIRB and non-AIRB (including ML) models. Established guidelines for algorithm selection, search strategies (e.g., grid/random/Bayesian), and performance metrics to align with regulatory expectations and internal model risk governance.
Developed and institutionalized Model Interpretability Policy Standards, integrating explainability frameworks (e.g., SHAP, LIME, feature importance) into model development lifecycle. Ensured that all models—especially complex ones like ensemble and neural networks—met transparency requirements for internal validation, audit, and regulatory review.
Temporal shapelets for intelligent feature engineering, extracting discriminative time-series patterns to enhance model performance and interpretability. Applied shapelet-based transformations to capture latent temporal dynamics in credit behavior, enabling more robust risk segmentation and predictive accuracy.
Manager
PWC
08.2021 - 09.2022
Developed AIRB - EAD model for 'Warehouse Facilities'
Developing Rating model for US Pension Funds using qualitative and quantitative factors.