Algorithmic Lending & Model Governance
Credit-scoring models on the German Credit (Statlog) dataset, with SHAP explainability and a fairness/bias audit mapped to Indian digital governance frameworks.
I enjoy making things. Here are a selection of projects that I have worked on over the years.
Credit-scoring models on the German Credit (Statlog) dataset, with SHAP explainability and a fairness/bias audit mapped to Indian digital governance frameworks.
A Basel-style expected credit loss engine combining calibrated PD with LGD and EAD, implementing IFRS-9 three-stage impairment logic.
Unsupervised anomaly detection on USD-INR exchange-rate data using Isolation Forest and rolling-volatility features.
An NLP pipeline over the 8+ GB CFPB Consumer Complaint Database that classifies complaints and separates templated filing volume from genuine consumer harm.
MS project on risk-neutral valuation and delta-hedging of vanilla and binary European options, including Greeks, implied volatility, and a simulated volatility surface.