Hackathons
The hackathon projects I’ve participated in are listed below.
Seattle: 2026
Cadence
Participated on June 19-20, 2026 (Link to Devpost Submission, Link to GitHub Repository, Link to Presentation)
Cadence, the winner of the Best Voice Award at the Changbal Innovation Hack 2026, is a voice-to-insight web platform designed to streamline personal journaling, therapy sessions, and coaching records by automatically synthesizing unstructured audio into transcripts, summaries, and actionable checklists. Developed in collaboration with John Lee, an undergraduate biomedical engineering student from Georgia Tech, the responsive web platform utilizes real-time audio capture pipelines, speech-to-text models for cross-lingual transcription, and dynamic backend text processing models to reduce administrative burnout and lower the activation energy for daily personal reflection. The core multi-role architecture—built using Lovable and BizCrush—features three distinct operational modes (Journaling, Therapist, and Coach) optimized to bridge communication barriers.
Should I Open Here?
Participated on May 29-30, 2026 (Link to GitHub repository)
This application is an AI-powered business location analyzer that evaluates the viability of a specific address for a given business type. It coordinates a FastAPI backend and a vanilla JavaScript frontend to scrape competitor data and hourly traffic congestion metrics via Apify, processes these signals deterministically to compute a market saturation score, and uses Anthropic’s Claude to generate a structured markdown report. The finalized report is then automatically uploaded to Box, returning a shareable file link and an in-page preview to the user. I co-developed this application alongside Madan Mohan Chundari, Ritunjay Murali, and Ayah Zaheraldeen — the same team as the one that developed Prospectus (shown below) — for the Cascadia AI Hackathon 2026.
Prospectus
Participated on May 21, 2026 (Link to GitHub repository)
At the HackwithSeattle hackathon on May 21, 2026, run by Devnovate, I co-developed Prospectus alongside Madan Mohan Chundari, Ritunjay Murali, and Ayah Zaheraldeen. Our project is an open-source financial research assistant that parses SEC EDGAR 10-K annual reports into a vector database to deliver cited, real-time responses to complex financial queries via an iMessage conversational interface. Together, we built the entire end-to-end pipeline—encompassing automated scraping, text chunking, and semantic LLM retrieval—within a strict four-hour time limit, ultimately earning 1st place overall out of all competing teams.
Berkeley: 2022
Transit and Housing in California
Participated on November 11-13, 2022 (Link)
This article was made within 3 days for the Fall 2022 UC Berkeley Datathon for Social Good and won second place in the Urban Studies track. For this article, my Datathon team (consisting of Gain Boonavich, Anita Ding, Yixin Huang, and myself) used Python to perform linear regression and create data visualizations to dive into the relationship between the amount of investment in transit and the number of housing units in Californian counties.