I built a credit underwriting and macro stress decision engine that combines SQL-based transaction analysis, a calibrated LightGBM model, and Monte Carlo simulations to assess portfolio risk.
I deployed the engine as an interactive Streamlit dashboard, translating risk distributions into decision levers. I also quantified Value-at-Risk and Expected Shortfall under simulated macroeconomic and interest rate shocks.
At Escon Info Systems, I extracted and analyzed structured datasets with SQL and Python, automated recurring KPI reports, and investigated data inconsistencies through root-cause analysis.
With the Uttar Pradesh Power BI Club, I design and showcase dashboards and dimensional models, and facilitate workshops on visual storytelling and dashboard performance.
