Analyzing News Sentiment on YouTube and Snapchat’s Impact on Youth Mental Health
This project evaluated NLP sentiment classification models (Logistic Regression, LSTM, and fine-tuned BERT) to analyze news coverage of social media’s impact on youth mental health. My partner and I designed zero-shot prompts to integrate multiple LLM APIs (OpenAI, Gemini, Grok) for automated data filtering and sentiment annotation, then we validated the LLM outputs with human-validated ground truth labels. We also performed statistical analysis to compare the LLMs to each other and the human labels using Pandas, NumPy, and scikit-learn.
NYC Motor Vehicle Good Samaritan Website
This full-stack web application enables secure public submission of eyewitness reports for NYC motor vehicle crashes, translating real-world user needs into a reliable, user-centered workflow. I led a four-person team to design, build, and refine core product features, emphasizing data quality, authentication, and iterative testing throughout development, while contributing the core backend features of the skeleton code, comments and the user collection.
Software Engineering Internship at Hamilton Square Partners Management, LP
Hamilton Square is a New York area-based investment firm focused on the biopharmaceutical industry with >$300m AUM. I engaged in confidential software development and trading strategy; reference contact available upon request.
I also built the fund’s public-facing website with the Front-End framework Bootstrap