Portfolio
AI Engineer & MS Data Science student at Boston University — building at the intersection of machine learning, probabilistic methods, and responsible AI.
I'm an AI Engineer and researcher pursuing my MS in Data Science at Boston University. My work spans probabilistic machine learning, computer vision, and the responsible deployment of large language models — with a particular interest in how we design systems that are both technically rigorous and safe for real-world use.
I enjoy working across the full arc of a project — from formulating research questions and designing experiments to building end-to-end pipelines and shipping usable tools. I've collaborated with psychology researchers, climate-tech teams, and academic labs, which has shaped how I think about the gap between a model's performance on a benchmark and its impact in practice.
When I'm not writing code or reading papers, I'm usually thinking about how AI intersects with user vulnerability, social impact, and ethical design.
A novel sampler that adds lightweight "scout" evaluations before each Hamiltonian trajectory, steering momentum away from low-probability regions before committing to a full trajectory. Activates only when the chain is struggling — zero overhead otherwise.
A reproducible evaluation framework for safety-critical LLM behavior across vulnerable user groups — focusing on content safety, minor exposure controls, and elder-targeted scam detection in Gemini and Llama.
A computer vision pipeline that helps psychology researchers automate behavioral coding of child anxiety in video experiments — replacing days of manual annotation with a minutes-long automated workflow integrated into a web app.
Uncovering relationships between farmer age demographics and agricultural production across 3,000+ U.S. counties from 1997 onwards, with inflation-adjusted time series analysis.
An AI-powered web platform for climate-tech researchers to explore and rank crops for cellular agriculture, using multi-criteria scoring on environmental impact, regulatory feasibility, and commercial potential.
Open to research collaborations, engineering roles, and interesting conversations about ML, AI safety, and probabilistic methods.