The unglamorous half.

I build the layer between a raw feed and a decision. Most of my work has been the unglamorous half of that: getting a warehouse to agree with itself, writing the validation that catches schema drift before a dashboard quietly starts lying, and defining a metric once so four teams stop defining it five ways.

Three years of that ran across BI and data engineering — Power BI and DAX in front, GCP, BigQuery, Airflow, and dbt behind. Then a Master's in Data Science at Buffalo, and a capstone with Nissha Medical Technologies putting computer vision on a production line rather than in a notebook.

The NYC subway pipeline is the project I keep returning to, because the interesting part is not the model. The MTA never publishes an arrival. Everything a rider actually wants to know sits downstream of an event that has to be inferred, defended, and tested — which is most data work, with the difficulty made visible.

Beyond the résumé: Black belt in Okinawa Shorin Ryu karate · ACM Python mentor · Lead guitarist.