Seattle, WA · U.S. Citizen
Data Scientist · Applied AI · Analytics Engineering
I build validated data pipelines, statistical and machine learning workflows, and LLM evaluation systems that turn messy inputs into reproducible decisions.
Selected work
I wanted a small project that went beyond a notebook: a message goes through an API, the prediction is saved, and uncertain cases are left for a reviewer. It runs locally with FastAPI, Streamlit, and SQLite.
I built a preference-classification pipeline around 135K public Arena votes, then tested grouped splits, soft labels, pairwise scoring, and a frozen temporal holdout.
I joined hourly prices, generation, weather, and archived forecast runs for DE-LU, France, and Austria. The raw files stay available, while the curated panel is used to compare forecast errors with price movement.
I built a stock panel with historical index membership, walk-forward evaluation, transaction costs, placebo tests, and a locked holdout. The result is exploratory evidence, not a claim of persistent alpha.
Experience
Early-Stage Data & AI Startup
Built Python and SQL workflows, a three-layer data lake pattern, model-selection pipelines, RAG evaluation, and repeatable technical handoff assets.
Feelie
Standardized 500K+ conversation records and connected DistilBERT emotion classification, LLM response generation, and rubric-based output review.
UW Engineering Services
Built validated ETL and SQL reporting workflows, improved query performance by 40%, and automated 200+ hours of annual reporting work.
UW Department of Civil Engineering
Processed 1.2M+ sensor records per month with timestamp alignment, missing-value treatment, noise filtering, and feature engineering; built congestion-monitoring datasets, applied clustering to identify recurring traffic patterns and high-risk locations, and evaluated ARIMA and LSTM forecasting approaches for short-term traffic prediction.
Capabilities
I use fixed effects, regression, clustering, cross-validation, and time-series models when the question needs an interpretable comparison.
scikit-learn · statsmodels · XGBoost · ARIMA · LSTM
I build small evaluation sets, write labels that a reviewer can apply, and check where a model is incomplete, ungrounded, or simply wrong.
PyTorch · scikit-learn · LangChain · OpenAI API
I use Python and SQL to move raw files into tables that can be checked, rerun, and explained to someone who did not write the pipeline.
Python · SQL · pandas · Parquet · AWS · Tableau
Education
University of Washington · Seattle, WA
Scientific computing, machine learning for large-scale data, database systems, probability, and statistical modeling.
University of Washington · Seattle, WA
Data structures, algorithms, numerical analysis, statistical methods, and computational mathematics.