FEATURED PROJECTS
SQL: E-Commerce Analytics
Overview:
Analyzed a large-scale e-commerce dataset (250K+ transactions from 2020–2023) to uncover customer behavior patterns, product category trends, return risks, and opportunities to improve revenue and profitability.
Python: Fraud Detection & Risk Clustering
Overview:
This project focused on identifying fraudulent online transactions using machine learning models such as Random Forest and XGBoost. We engineered features from transactional and identity data, applied SMOTE to address class imbalance, and clustered transactions using GMM for risk profiling. The Random Forest model achieved 93.4% precision and minimized false positives to 0.2%, supporting smarter fraud detection and decision-making.
Tableau: Enchantments Permit Strategy Dashboard
Overview:
Our team collaborated with Northwest Treks, a Seattle-based adventure outfitter, to design a Tableau dashboard supporting their Core Zone permit strategy in the Enchantments.
Leveraging 2024 lottery data and a decade of historical weather trends, we built an interactive tool that:
Visualizes permit odds by date, zone, and group size
Recommends optimal windows based on rainfall and temperature patterns
Improves permit success rates from 2.6% to 22.17%
