I build machine learning systems — from the first notebook experiment to the production pipeline serving millions of predictions a day.
Focus areas
I'm a senior data scientist who works across the full lifecycle — framing the business question, building the model, and shipping it behind an API that survives real traffic. I care as much about latency and monitoring as I do about offline F1 scores.
A model that never leaves the notebook hasn't shipped anything. I default to simple baselines, instrument everything, and only reach for a bigger model when the data and the metrics ask for one.
Streaming gradient-boosted model scoring transactions in under 80ms, replacing a rules-only system.
Hierarchical time-series model forecasting SKU-level demand across 600+ retail locations.
Fine-tuned transformer classifier that routes inbound support tickets to the right queue automatically.
Propensity model paired with an experimentation framework to test retention offers against predicted churn risk.
Lead ML systems for the growth and risk teams. Own the model lifecycle from experimentation through production monitoring.
Built forecasting and pricing models for a multi-region retail chain. Partnered directly with supply-chain planners.
Started in analytics, transitioned into building the company's first production ML model for route optimization.
Amazon Web Services
2023
2021
Kaggle
2022
University of Washington
2016
Whether it's a model that needs to ship, a pipeline that needs to scale, or a metric nobody trusts yet — let's talk.