Great Lakes Buoy Forecasting
Python · Pandas · NumPy · scikit-learn · PostgreSQL · SQLAlchemy · Requests · rsync · Cron |
- Built an **hourly** end-to-end pipeline that ingests live NOAA buoy feeds, cleans them with Pandas, and stores snapshots + aggregates in PostgreSQL.
- Retrains lake-specific Ridge models daily, generating 28-day water-temperature forecasts with persisted coefficients.
- Automated an evaluation layer (MAE / RMSE) that writes accuracy metrics back to the DB for Tableau dashboards.
- Secured everything on an Ubuntu VM (UFW, SSL Postgres, cron, health-checks, Git + rsync sync).
Personal Blog Platform
Node.js · Express · SQLite · EJS · CSS · JavaScript |
- Full-stack web app with server-side rendering and persistent storage in SQLite.
- Implemented secure authentication with bcrypt; managed sessions with
express-session. - Responsive UI written in vanilla CSS—optimised for simplicity & speed.
- Deployed on a Linux VM with environment-based config and automated backups.