In February 2024, at Stanford TreeHacks, my team developed Synthesis, a groundbreaking news platform aimed at combating political polarization and information overload. Leveraging cutting-edge AI technology, Synthesis aggregates over 60,000 articles daily from 100+ sources, clustering them by topic similarity to provide unbiased and comprehensive coverage. Through innovative features such as Gists, Syntheses, and a Recommender Algorithm, users can access personalized, high-quality news content that promotes understanding and engagement. I led the development of Synthesis's AI News Agent, utilizing advanced techniques in text extraction, semantic similarity, and clustering to create a scalable and cost-efficient system. We won Best AI Hack, sponsored by Coframe.

In fall 2023, I co-led product and engineering development for a team of 4 software engineers to create a web platform for waste management in Fiji using React. Created project roadmap, designed initial product wireframes, coordinated meetings with our client, and conducted code reviews through pull requests.

In summer 2023, I developed Classiq, a Harvard course planner and scheduler. I found a need for such a service due to how slow and ineffective My.Harvard, Harvard's main course search tool, was. Using Classiq, students can search for relevant classes quickly through fuzzy search, sorting, and filtering. Students can view their classes on a schedule matrix and export the classes that they want into their Apple and Google calendars. Classiq currently has 2K+ unique users and 17K+ page visits with an abundance of positive reviews.

In my freshman spring semester at Harvard, I led a group of fellow software engineering students to develop the Visual Analytics portal for City of Boston. The Visual Analytics portal compiles the public data from the City of Boston's Checkbook Explorer into a comprehensive data visualization tool that allows anyone to become better informed on spending by the City of Boston. This tool improves over the existing site which simply provides CSV data for users, potentially being inaccessible to users who are not experts in data analysis tools like Excel. Working with the City of Boston, we developed a React/Django-powered web app where users can view total spending across different years, view departmental spending, analyze how spending has changed over time, and peruse vendor-specific spending. Through this portal, the City of Boston's spending is more transparent and accessible to Boston citizens.

In my freshman fall semester at Harvard, I collaborated with fellow social tech entrepreneurs to develop a web platform for OkaySo. OkaySo is a nonprofit initiative that provides a safe space for young adults to anonymously ask questions to experts regarding mental health, identity, relationships, and more. These are topics that many young adults have questions about as they grow older but do not feel comfortable doing so publicly and need authentic, educated advice. However, we found that current online solutions, including OkaySo, were not as readily accessible to both inquirers and experts. Working with OkaySo, we created a React/Node.js-powered web app where experts can efficiently respond to user inquiries and facilitate seamless, secure, message-based conversations with those whom they are helping. As part of our efforts, we were able to substantially increase the team's user base and user satisfaction with expert replies, bringing us one step closer to improving health education.