Tech Stack
A quick snapshot of the tools I use most.
Experience
My professional and academic journey
- Selected as 1 of 4 interns from 300+ applicants to build an AI-powered fundraising platform using Flask and Python, projected to drive $10M+ in gifts by connecting philanthropists with Duke research initiatives
- Increased donor-research match accuracy by 40% by embedding donor interests with Sentence Transformers and mapping 2K+ alumni relationships in Neo4j to surface high-value connections for targeted outreach
- Accelerate donor data acquisition by 95% by building a Selenium pipeline to scrape 13K+ records and enriching profiles with DukeGPT insights spanning board roles, affiliations, and career history
- Developing Duke's largest mentorship program through a Python-based matching system that pairs mentors and freshmen by shared interests and backgrounds, scaling to support 12.5K+ participants
- Improving match accuracy and reducing manual processing time by 80% by cleaning and standardizing 100K+ survey responses across multiple semesters with Pandas
- Collaborated with fellow first-gen students in Duke's Spring Forward Program to deliver financial forecasting tools; Optimized AI-driven stock forecasts using Python and OpenAI API, improving accuracy by 35% through data tuning
- Implemented a financial data pipeline using Python, Pandas, and RESTful APIs to process 1M+ daily stock price records across 5,000+ companies, and analyzed large-scale stock price data for real-time market insights
- Automated Python web-scraping pipelines with BeautifulSoup to collect and process financial disclosures, cutting manual data gathering by 9+ hours weekly and enabling large-scale political finance analysis
- Structured and validated datasets of legislative financial records with 98% accuracy in Excel, supporting reliable research on lawmakers' business affiliations and investment activity
Leadership
Expanding Duke's largest pre-professional tech org by hosting resume workshops and professional events; generated 150K+ social media impressions and boosted event attendance by 30% YTD
Mentoring first-gen/low-income students across multiple cohorts, easing academic and social transitions and driving a ~25% increase in campus resource use
Projects
🏆 2x Hackathon WinnerAward-winning projects and personal builds
Built an AI voice agent using Flask and Retell AI to bridge users to 800+ verified aid programs, improving access to housing, food, and energy assistance in underrepresented rural communities.
Enhanced the system with multilingual and SMS support, delivering real-time voice assistance at ~60%lower cost than human operations.
Built a cross-platform mobile app to manage poker sessions, automating buy-ins and earnings across 30+ games, reducing settlement time by 90%
Implemented QR-based game/user encoding with Expo Camera cutting setup time from 2 min → 10 secs and real-time sync with Firebase Firestore for game state and earnings
Led development of a disaster-response app that connects victims with volunteers via real-time geolocation, enabling faster mobilization and more effective resource allocation during crises.
Enabled the mobile frontend to process 500+ live help requests during testing by developing and deploying backend services with Node.js, Express, and PostgreSQL.
Built a modern, responsive personal website featuring real-time Spotify integration with my listening stats, smooth scrolling animations, and a unique gradient design system.
Integrated Spotify Web API to dynamically display my top artists and tracks with real-time updates.
Hobbies & Interests
What I do outside of coding
View my photo collection from travels and everyday moments.
Explore my travel map and photos from places I've visited.