Projects

Collaborated with Komatsu Corporation to analyze large-scale mining data on Databricks. Identified critical outliers, applied t-SNE clustering, and built two interactive Plotly dashboards.

AI-powered crypto portfolio advisor combining neural collaborative filtering and real-time market regime analysis. Personalized token recommendations backed by 200k+ wallet interactions and dynamic risk optimization.

Custom-built dashboard analyzing Salesforce SEC filings data using PowerBI and Python. Provides interactive views into revenue, expenses, and growth trends.

A comprehensive geospatial and statistical analysis of NYC Taxi trip data, combining backend statistical modeling with a production-ready interactive dashboard.

Stress and emotional detection using the WESAD dataset with consumer-grade and medical-grade wearable sensors (Empatica E4, RespiBAN) to identify physiological stress markers.

Engineered features and validated ML models (XGBoost, Random Forest) for patient diagnosis. Achieved >99% recall and precision with the top-performing model.