Current Research

My work revolves around understanding how developers interact with AI-assisted tools and what we can do to better support them. I'm currently exploring techniques to improve interaction with AI Assistants and benchmarks for AI-generated code.

Research Experience

Aug 2025 · Present

Graduate Research Assistant

George Mason University - Schar School of Policy and Government

• Fine-tuned Llama models for structured extraction of entities and attributes from generated text, replacing manual annotation across thousands of model outputs • Conducted large-scale computational analysis of historical scholarly discourse across multi-decade corpora using machine learning techniques • Developed and facilitated workshop materials introducing large language models to quantitative social science researchers, covering probabilistic behavior, generation parameters, and model output analysis

Aug 2022 · Present

Graduate Research Assistant

George Mason University - INSPIRED Lab

• Investigated ambiguous user prompts for text-to-SQL task, using hidden-state geometry to identify internal signals associated with downstream LLM behavior. • Modeled task difficulty in software issue resolution benchmarks using repository, patch, and prompt-derived features from a large scale dataset using ensemble methods and SHAP-based explainability • Explored automated prompt decomposition strategies for LLM-based code generation and built a VS Code extension prototype to improve generation accuracy • Led a mixed-methods qualitative study to understand developer information seeking behavior while interacting with AI-enabled assistants • Analyzed survey data to identify and address engagement barriers in technology for marginalized groups

May 2023 · Aug 2023

Research Intern

Microsoft Research

Developed and presented a prototype for eliciting mental models from software artifacts using LLMs

June 2022 · May 2023

NSF Research Trainee

George Mason University - Center for Advancing Systems Science and Bioengineering Innovation (CASBBI)

• Designed and conducted semi-structured stakeholder interviews to understand problems encountered by wheelchair users when navigating their surroundings • Developed a mobile application prototype based on findings from stakeholder interviews to assist wheelchair users navigate around the campus • Presented findings to stakeholders at the CASBBI Annual Retreat

Let's Connect

Feel free to shoot me an email if you'd like to discuss research, collaborate, or have any questions about my work!