NOAA Fishing Survey Data Project
This project demonstrates an ETL pipeline that processes NOAA's fishing survey data, then makes it available for analysis through an interactive dashboard.
Welcome to my portfolio site, where you can explore my data-related projects.
This project demonstrates an ETL pipeline that processes NOAA's fishing survey data, then makes it available for analysis through an interactive dashboard.
This project ingests YouTube video data related to fishing, stores it in MongoDB, and provides visualizations through Metabase for analysis.
Serverless monitoring system that runs YOLOv8 object detection on a live NJ beach stream every 30 minutes via GitHub Actions. Uploads annotated detection clips to Cloudflare R2 and sends Gmail email alerts when car or person thresholds are exceeded.
Interactive single-page application for mortgage and financial planning. Compare monthly payments and total cost across 15, 20, and 30-year loan terms with per-term interest rates, visualize amortization schedules, and explore how loan size and property taxes impact debt-to-income ratios.
Personal saltwater fishing analytics pipeline that ingests Garmin GPS activity files, automatically enriches trips with weather, tide, water temperature, and solar data from public APIs, and analyzes catch patterns across environmental variables.
Interactive browser-based calculator for Shapeoko CNC routers providing instant feed rate, plunge rate, depth of cut, and chipload recommendations based on tool type, material, and spindle speed. Zero dependencies — runs directly in any browser.
Full-stack discrete event simulation of domino games with pluggable player strategies and Monte Carlo batch orchestration. A completed React + FastAPI application lets you run single games or batch simulations, compare strategy win rates with confidence intervals, and visualize Monte Carlo convergence — containerized with Docker.
Retrieval-augmented generation system for industrial equipment documentation. Upload PDF manuals through a Streamlit interface and ask natural language questions powered by OpenAI embeddings and GPT-4o-mini via a RAGAnything framework.