I am an Applied Scientist specializing in Sequential Decision Analytics (SDA) and reproducible optimization. My work focuses on designing and deploying online decision policies under uncertainty that deliver measurable economic impact.
Currently, I serve as the Leasing Reporting Manager at Phoenix Tower International, where I bridge the gap between data engineering and decision science. I leverage the Warren B. Powell framework (specifically PFA and DLA classes) to transform massive datasets into optimal real-time decisions, ensuring 100% reproducibility through Nix, Docker, and GitHub Actions.
Professional Journey
My expertise lies in moving beyond deterministic batch optimization toward simulation-driven decision-making. I have a proven track record of building production-grade simulation environments—including Monte Carlo and Discrete Event Simulations—to solve complex operational challenges.
I am a strong advocate for open-source R development and have authored several packages:
- {tidyvalidate}: A production-grade validation layer for identifying business logic errors.
- {biblegatewayr}: A tool for programmatic web-scraping of Bible verses.
- {corrcat}: A package for exploring associations between categorical variables.
I also contribute to the broader ecosystem, including documentation for the {data.table} package regarding high-performance join operations.
Featured Project: NYC Taxi Earnings Optimization
I am currently developing a sequential decision support system designed to maximize NYC taxi drivers’ net hourly earnings. Rather than a standard prediction model, this project implements:
- Policy Evaluation: Testing different driver strategies through high-fidelity simulations.
- Lookahead Approximations: Using future-state simulations to inform current trip acceptance.
- Reproducible MLOps: A fully containerized environment using Nix + Docker and automated deployment pipelines.
The goal is to move from probabilistic predictions to profit-driven thresholds, demonstrating a 51.5% potential uplift in expected earnings through optimized decision policies.
Skills & Education
I hold a B.S. in Industrial Engineering from INTEC (GPA 3.91/4.0) and am currently pursuing a Master of Data Science (expected 2026). My technical stack is built for scale and precision:
- Decision Science: Powell Framework (PFA/DLA), Monte Carlo Simulation, Discrete Event Simulation.
- Core Stack: R ({data.table}, {sf}, {tidymodels}, {pins}) and Python (Polars, Flask, SQLAlchemy).
- Engineering & MLOps: Nix, Docker, DuckDB, GitHub Actions, and spatial data engineering.