Big Waves

Solo project · big-waves.vercel.app

A React tracker for the Mega Man Star Force battlecard collection. Browse the full SF3 card library, mark cards obtained, and track collection progress across game versions, all persisted client-side.

Rebuilt the production asset pipeline after discovering card art (a ~790MB scanned image library) was silently dropped from the production build, then cross-referenced hundreds of untitled scans against their Japanese card names to correctly attribute art across the full card set. Deployed on Vercel with Web Analytics wired in for private traffic insight.

MarketZone

3 contributors · API + React storefront, both live on Vercel

Production REST API for a full e-commerce marketplace. Handles user registration and JWT authentication, product listings, shopping cart, Stripe payment processing, order management with billing and shipping addresses, reward points, and automated order confirmation emails.

Paired with a React storefront covering login, registration, account management, product posting and a cart-to-payment-to-confirmation checkout flow using Stripe Elements.

SingWithFriends

Host/guest roles · Custom auth middleware · v0.1.6

A networked, multi-user music control system built in C# and .NET 10. An embedded ASP.NET Core Web API (Kestrel) runs inside a WinForms desktop app: friends join via room code from any browser on the network, upload songs from their phone, and control playback in real time.

Includes custom API key and guest token authentication middleware, request logging middleware, file upload validation, Swagger docs, and a full GitHub Actions CI/CD pipeline with semantic versioned releases.

Grid Pathfinding Benchmark

Group project · Master's coursework, Fall 2026 · JPS and JPS+ next

A C# harness for benchmarking grid pathfinding algorithms on 8-connected maps. Implements A* with an octile-distance heuristic, a custom binary min-heap, and flat-array bookkeeping for speed, behind a common pathfinder interface so Jump Point Search variants can be compared head to head.

Parses plain-text and Moving AI benchmark maps, runs randomized start/goal queries with warm-up and averaged timings, and reports path cost, nodes expanded and runtime. An xUnit suite checks path validity, diagonal movement rules and correctness against a reference Dijkstra search.

MAD Delivery App

Team project · Demo available

Cross-platform mobile delivery app built with Flutter and Firebase. Dual user and driver role architecture, real-time order tracking, nested Firestore review system, and Firebase Authentication. Built as a collaborative team project with a live demo.