🚀 I build tools with intention — fast, scalable, and insight-driven. Whether it's frontend, backend, automation, or shipping under pressure, I deliver.
Live at trainpace.com
Frontend (React + TypeScript + Vite)
- GPX-based elevation analysis with interactive visualizations
- Pace + fuel planners with real-time adjustments
- Riegel predictor & smart race advisory
- Firebase Auth integration with domain-restricted flows
- Branch-aware deployments via Vercel
Backend
- Node.js API
- Features: elevation analysis, GPX parsing, fuel plan refinement
- Firebase integration for auth + Firestore persistence
- Modular endpoint design for feature expansion
Learning about incentive engines through manual UX observation
- No scraping, just strategic curiosity
- Case Study: Uber Eats 21-Day Engagement Mapping
Built to serve the runner I became. Insight-first tools for real-world pacing.
Tech Stack:
- Frontend: React, TypeScript, Vite, TailwindCSS, React Router
- Backend: Node.js, Express, Firebase Auth, Firestore
- Deployment: Vercel (branch-aware, domain-restricted auth)
- Testing: Playwright with full CI/CD automation
Key Features:
- GPX upload & elevation profile analysis
- Dynamic pace planning with terrain awareness
- Fuel strategy calculator with real-time refinement
- Riegel race time predictor
- User authentication with Firebase + protected routes
Repos:
Behavioral experiment mapping engagement escalation on Uber Eats.
No code. No scraping. Just sharp systems thinking and documentation.
I ship. I analyze. I mentor. I automate.
Whether it's shipping product features, reverse-engineering patterns, or debugging broken systems—I do it all with clarity and calm.
Core Strengths:
- Building fast, maintainable SPAs with strong separation of concerns
- API design with modular, scalable endpoints
- Firebase Auth integration across frontend/backend
- CI/CD automation with GitHub Actions + Playwright
- Real-world problem-solving for runners and engineers alike
🧵 Still Building. Still Learning.
Next up? AI race prediction and user coaching agents using MCP and Firebase.