Projects | Daniel Staggs

I have a significant number of completed projects awaiting publication, which will be made available on this platform in the near future.

Diet/Bulk Calorie Analyzer

• Problem solved: Forecast how much weight you should gain or lose during a bulk /cut cycle and compare it to what actually happened.
• Data sources: Apple Watch (Health CSV), Cronometer nutrition logs, Hevy workout exports.
• How it works: Pandas cleans & joins the CSVs → daily energy balance model → Matplotlib charts predicted vs actual weight-curve with error bands.
• Outcome: Gives athletes an instant sanity-check on their diet without manual spreadsheets.

Full Source →

Fantasy Football Analyzer (v1)

• Goal: Win waiver wire and trades by turning raw ESPN data into actionable scores.
• Stack: Python / Flask UI, ESPN private API, simple linear-reg model (scikit-learn) for rest-of-season value.
• Features: Waiver priority list, trade-value calculator, weekly start/sit optimiser.
• v2 roadmap: Live API polling, LLM-generated matchup notes, and containerised deployment.

Ubuntu Server Homelab

• Why: Cheap on-prem sandbox for automation and privacy-focused services.
• Infra: Ubuntu 22.04 LTS on mini-PC, accessed via SSH and Tailscale VPN; Docker + systemd for services.
• Workloads: Python scrapers, static-site hosting, uptime-kuma monitoring, and a locally-served 8-bit LLM during R&D.
• Ops: Automated Off-site backups (rclone → Wasabi), Nginx reverse-proxy, cron-based log rotation.

Gas Tracker PWA

Offline-capable web app that logs fill-ups, stores data in localStorage, and computes MPG & cost-per-mile.

Deep Dive →

Fantasy Football Data Hub

A fully automated pipeline that collects multiple seasons of NFL data via nfl-data-py, runs custom analyses (VORP, consistency ratings, matchup and tier calculators) using Python and GitHub Actions, and serves interactive charts to give data‑driven insights for drafting, weekly matchups and waiver pickups.

GitHub Repo →

CFB Game Data Model AI

College-football Game Prediction Data Model Project: With a live site and open code. This is actively evolving; the deep-dive page will track the approach and results.