The site you're exploring right now! Built from the ground up during summer 2025, this project served as a side project to refine my front-end development skills. This platform is a space to showcase some of my work and personal experiences. Every element—from layout to styling—was designed and coded by me as part of the learning process.
A final project for Harvard's CS50 course, Veritas & Varieties is a web application designed to simplify and automate event discovery across the Greater Boston area. Our goal is to provide a website for users to discover and plan their participation in local events without manually going through newsletters or other resources.
As part of Harvard's CS 2860 course, this research project explored the robustness and performance of differentially private consensus algorithms under realistic network conditions. Working with teammates, we analyzed the performance of Laplace and Gaussian differential privacy mechanisms, packet loss, communication noise, and topology-aware scaling methods.
Incoming PM intern for the ZIA team
I am a teaching fellow for CS 1200, which had around 60 students. Beyond weekly office hours and sections, I hosted review sessions for the final, test-solved exams, and graded weekly problem sets on algorithms.
I was a teaching fellow for CS 50, which had around 400 students. I led weekly sections with tailored lesson plans, hosted office hours to support student learning, and graded problem sets and final projects with detailed, constructive feedback aligned with course objectives.
I worked as a Computing Intern at Lawrence Livermore National Laboratory, contributing to Sina, an LLNL codebase for data querying, storage, and visualization. As part of the Livermore Lab Foundation Fellowship, I designed circuits and logic gates using BJT transistors and built a basic 16-bit CPU, ALU, and program counter in Hardware Design Language.
I was a research assistant for LISH, which is under Harvard Business School. I analyzed how accelerator programs assess startups by reviewing and categorizing longitudinal judge feedback. Using this analysis, we then annotated and structured feedback data to fine-tune a large language model (LLM) for research on startup evaluation frameworks.
As a software and analytics intern, I identified a gap in the final step of internal paper publishing where meta-repositories, a single point of access to find all components of a paper, were often overlooked. To help solve this problem, I developed "autometa," a Python-based LLM tool to auto-generate meta-repositories for reproducibility.
My first summer at PNNL, I worked to analyze CERF data (Capacity Expansion Regional Feasibility) to provide insights on power plant locations. Then, using Python, Plotly, and Dash, I built an interactive visual analytics app, providing researchers an ongoing resource.