Messy data in, decisions out.

I'm the person who takes messy data that's scattered across different systems and turns it into something a business can actually use to make decisions. Most of the work is in making it trustworthy: getting the numbers to agree with each other, setting up checks so they stay right, and defining a metric once so four teams stop defining it five ways.

I did that for about three years, first on a Google project through Smollan and then at Constituents AI. Then I came to Buffalo for a Master's in Data Science, and my capstone with Nissha Medical Technologies put computer vision on a real production line, not in a notebook.

These days I build the ML and AI on top of it too: a voice agent that answers as me, a search over FDA filings that refuses rather than guess, and the subway pipeline, which is the one I keep coming back to. The MTA never publishes when a train arrives, so everything a rider wants to know has to be worked out and tested. That's basically most data work, with the difficulty made visible.

Beyond the résumé: Black belt in Okinawa Shorin Ryu karate · Cooks a lot (not on this portfolio means in the kitchen).