Statistics & Data Science Major @ UCSB
Hi, and welcome to my portfolio website. Here you will find some of my relevant experience & other misc stuff I’ve been working on.
Container infrastructure for university-wide data science compute environments. Built a multi-arch CI/CD pipeline in Jenkins that builds ARM64 and AMD64 images in parallel and aggregates them into a single Docker Hub release.
Shipped a multi-architecture container image (x86/ARM) with automated manifest aggregation, published to Docker Hub.
A 24-hour hackathon build. Chains Deepgram, Gemini Vision, and Gemini Flash to turn live audio and video into real-time social context summaries — an accessibility tool for neurodivergent users.
Shipped a working demo in 24 hours.
An autonomous drone that runs a computer vision model on-device. Built from scratch: OS install, hardware wiring, and a YOLOv11 model for real-time ball and player detection for a soccer coaching assistant.
Backend for a real-time collaborative band app — rooms, synchronized playback, playlists, and shared sheet music. Included a scraping pipeline that generates PDF/WebP assets on cache miss.
Cut mean latency from 720ms to 96ms; 7.4× efficiency gain and 2.1× throughput improvement (4.3 → 9.2 req/s).
My personal site — a production full-stack app I maintain end-to-end. Includes a regression calculator that runs gradient descent on user-uploaded CSVs and a browser-side ONNX inference feature.
A YOLOv5 object detector that recognizes 12 Naruto hand signs from a webcam feed. Replaced an earlier Mediapipe + Random Forest pipeline that couldn't handle real-world variance.
Trained on a custom-annotated dataset with hue jittering and rotation augmentation; being ported to ONNX for in-browser inference.
Run workshops and office hours on computer vision, web hosting, and regression analysis for 30+ project teams. Started as an intern and grew into the director role.
the flip
Off the clock — music, photography, and experiments. Pick a room.