At Accenture S&C, I embedded predictive analytics into operational reporting for Nestle, tracking more than 15 KPIs and improving milestone forecast accuracy by 20%. I also integrated ML anomaly detection into the ERP-to-OMP pipeline, vetting over 500,000 records and reducing validation time by 30%.
At the University of New South Wales, I developed an end-to-end churn pipeline for more than 500 records and trained and evaluated Random Forest and XGBoost models. I engineered behavioral features and user segments to identify customer retention drivers.
In my Credit Card Anomaly Detection System project, I created a two-stage pipeline using autoencoders and LightGBM, reaching 0.71 recall at a 0.5% review budget. I also added SHAP explanations to surface the top three drivers for each fraud alert.
My other projects include a financial expense forecasting analysis across 2.4 million records and an investor intelligence platform using hybrid search and RAG. At IIT Delhi, I studied Mathematics & Computing and received a KVPY fellowship, along with national and state-level academic distinctions.

