Data & Business Analyst in Boston. I turn messy, multi-source data into reporting and recommendations people act on.
My best work starts with a number that doesn't add up and ends with a business decision that changed because of it. MS in Business Analytics from Northeastern University.
| Project | What it shows | Stack |
|---|---|---|
| Donor propensity model | Captured 37.7% of likely planned-gift donors in the top 1% tier, on a 0.1% positive class across 318K+ constituents | Python · XGBoost · LightGBM |
| Student retention dashboard | Who drops out and when to step in: ~80% dropout with no first-semester units passed vs ~10% with 6+ | Power BI |
| U.S. airfare trends | 11 interactive views of fares before and after the pandemic, in nominal and inflation-adjusted terms | Tableau · SQL |
| Ride fare prediction | Cut query latency from 4 min to 12 s on 690K+ records; XGBoost fares to $6.69 MAE | BigQuery · PySpark · Flask |
| Football match agent (in progress) | Plain-English questions answered by an XGBoost model, with Claude choosing the tool | Python · DuckDB · Claude API |
| Spotify listening behavior | Skip behavior and K-Means segmentation of ~150K streaming records | Python · scikit-learn |
Case studies for each, plus my work experience, are on my portfolio.
BI & visualization: Power BI (DAX, data modeling), Tableau, Excel (Power Query, Power Pivot, VBA)
Querying & data: SQL, Python (pandas, NumPy), R, Google BigQuery, PySpark, Alteryx
Analysis & ML: regression, classification, segmentation, XGBoost, LightGBM, scikit-learn, hypothesis testing