React.js web app allowing users to create Balatro mods using a Scratch-inspired interface.

• Developed a bespoke code generation algorithm that transforms visual code blocks into complete SMODS-compatible Lua files.
• Designed a block-based interface that abstracts branching IF statements, complex conditions, and random number generation.
• Managed open-source contributions through GitHub, resolving 200+ issues and 100+ merge conflicts.
• Built a CI/CD deployment pipeline to Cloudflare Pages.
• Scaled the platform to an average of ~1,000,000 requests per month.

Cross-platform desktop mod-development platform for Balatro with a Rust backend and modern TypeScript UI.

• Built and shipped a cross-platform desktop product with Tauri, Rust, React, and TypeScript.
• Designed Rust command and service layers with Serde and IPC patterns to reduce frontend complexity.
• Developed a scalable rules engine using graph-based modelling, schema validation, and code-generation pipelines.
• Implemented typed data contracts and structured validation to improve data integrity.
• Built reusable UI foundations with Radix UI and Tailwind CSS across 60+ components.
• Added automated Rust tests, regression checks, GitHub Actions release automation, and multi-platform packaging.
• Integrated PostHog analytics to track feature usage and support product prioritisation.

Next.js web app to track, index, and analyse roguelike games, allowing users to browse ratings and create lists.

• Designed a comprehensive metric system evaluating genre heuristics including narrative presence, synergy depth, and RNG reliance.
• Implemented secure authentication using JWT tokens and Bcrypt hashing.
• Engineered a crowd-sourced validation workflow where user contributions are vetted through community voting.
• Developed a dedicated email microservice for password recovery.
• Optimised SQLite read performance by ~300% through indexing and query optimisation.
• Built web scraping flows to aggregate pricing and review data from Steam, IGN, and Metacritic.

React Native mobile application designed to gamify the process of learning and progressing in Skateboarding.

• Developed a recommendation algorithm that dynamically suggested new tricks based on user heuristics.
• Scripted a calorie estimation algorithm based on skill level and completed tricks.
• Designed progression systems to gamify learning and encourage repeat practice.
• Deployed to the Google Play Store, reaching 1K+ downloads.
• Created for my dissertation, receiving a First Class Honours.