Software Developer · UCSB
CS student at UCSB who loves building things at the intersection of systems, AI, and the web. Driven by genuine curiosity about how things work under the hood.
Researching frequency-aware authenticated data structures that reduce proof sizes and computational overhead for blockchain infrastructure. Part of Professor Dahlia Malkhi's Foundations of Financial Technology lab, working primarily in Go.
Leading and supervising the web development intern team while managing deployment, maintenance, and administration of multiple sites and servers via FTP. Build interactive minisites and internal tooling for the games and data desks, monitor analytics and dependency health, and keep the stack running across departments.
A self-hosted photo manager with AI tagging and natural language search, running on a Raspberry Pi 5. Uses Azure Computer Vision for auto-tagging, AWS Rekognition face collections via IAM Roles Anywhere, a pgvector-powered Bedrock RAG chatbot, and Cloudflare Tunnel connectivity — all designed to keep your photos private.
Visit project →A learning management system that catches AI cheating with concrete evidence rather than probabilistic guesses. Invisible context-aware markers are embedded in PDFs via AltText; when a student's prompt matches above 75%, the system triggers an adaptive voice interview powered by Deepgram's Voice Agent with Gemini 2.5-based comprehension scoring.
View on GitHub →A budgeting web app for individuals and groups. Users upload bank transaction CSVs for automatic categorization, sync structured budget data back to Google Sheets, and collaborate through Google OAuth-backed accounts.
Visit project →Built React/TypeScript interfaces for 3D model rendering applications. Optimized performance with the Context API by only rendering dynamic models when in the user's line of sight, resolved raycasting selection bugs so only the first component in view is affected, and implemented a CI pipeline to standardize formatting and automate builds.
Built a front-end CI workflow with Storybook and Chromatic for visual regression testing and collaborative UI review, letting the team catch visual bugs before merges. Shipped responsive, accessible interfaces across platforms; participated in weekly sprints and 1:1 syncs to unblock teammates.