I'm a current Harvard sophomore studying Computer Science and Government with a secondary in Statistics.
I'm interested in web development, AI/ML, advances in LLMs/GenAI, finance/trading, STEM education, and the important questions asked in AI ethics.
I'm currently a project lead for HUMIC, Harvard's premier undergrad ML club; a course assistant for ECON50, a popular data science and public policy course; and an upcoming teaching fellow for CS50, Harvard's most popular CS course.
Favorite Technologies:
Languages:
Frameworks:
Libraries:
Tools:
Some Personal Projects:
Quaizar
Built a quizbowl question classifier/LLM generator and real-time competition site for high school competitors. Implemented a frontend question reader and score tracker via SvelteKit, TypeScript, HTML/CSS and real-time backend with PocketBase, SQLite3 and a question generator via OpenAI’s GPT-4 LLM API.
Panopticon
Built a web scraper and lecture caption summarizer for Harvard students’ online lectures. Implemented via Python and its Selenium, Streamlit, and OpenAI libraries for headless browser scraping, frontend hosting, and large language models API calls.
HUMIC Projects I've Led/Built for:
VeritasGPT
Leading a team to build an LLM-based chatbot to serve as the ultimate assistant for Harvard students. Using LLM agent and retrieval-assisted generation (RAG) techniques to extend LLM capabilities in Python and React.
Fabulous Fables
Developed a multi-modal storybook generator that creates text and image pages of a children's storybook. Built in Python and React using OpenAI libraries and GPT-3.5/Dall-E models.
sophomore studying Computer Science and Government with a secondary in Statistics.
Built a quizbowl question classifier/LLM generator and real-time competition site for high school competitors. Implemented a frontend question reader and score tracker via SvelteKit, TypeScript, HTML/CSS and real-time backend with PocketBase, SQLite3 and a question generator via OpenAI’s GPT-4 LLM API.
Built a web scraper and lecture caption summarizer for Harvard students’ online lectures. Implemented via Python and its Selenium, Streamlit, and OpenAI libraries for headless browser scraping, frontend hosting, and large language models API calls.
Developed a multi-modal storybook generator that creates text and image pages of a children's storybook. Built in Python and React using OpenAI libraries and GPT-3.5/Dall-E models.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent