Building intelligent and intuitive software.
Full Stack Development Intern @ AIM-HI · Irvine, CA
A few things I've built recently.
Agentic RAG system for document Q&A with citation tracking. Chat with your documents across multiple AI models, semantic search via vector embeddings, and automatic quiz/flashcard generation.
Renewable energy site-placement tool. Place renewable energy sources on a map and get a real-time sustainability score.
YOLO-based object detection to track players, referees, and the ball in soccer match footage.
1st place, supervised learning track. Combined multiple modeling approaches to achieve 97% accuracy.
Where I've worked and what I've shipped.
Shipping generative AI features to production, including a video generation pipeline (Runway + Google Veo 3.1) used by 20+ users, a 5-stage image-to-video pipeline with real-time SSE progress tracking, and LinkedIn analytics REST endpoints.
Benchmarked four ML architectures for embedded deployment, achieving 99.0% accuracy and 0.10ms latency, 97.5% below the real-time threshold; collaborated with Electrical Engineering researchers on deployment decisions.
Built a real-time React/TypeScript code editor for trading script validation, optimized dashboard components, and cut API response time 40% via async workers and Flask route optimization.
Technologies I work with.