Projects
These are independent projects — not extensions of my job titles. Some started because I had a problem I wanted to solve. Others started because I became curious about a system and wanted to understand how far I could take it.
They range from a shipped consumer app to trading automation, trust-and-safety tooling, operational dashboards and a Spotify library audit. What matters to me isn't that they all belong to the same category. They don't. What they show is how I approach unfamiliar problems: understand the system, define the rules, work out what can go wrong, build something, test it and keep questioning the result.
Built with AI-assisted development
Across these projects, I define the problem, write the specification, establish constraints and failure conditions, direct the implementation, review outputs, test behaviour and iterate. AI is part of my build process — not a substitute for deciding what should be built or whether the result can be trusted.
An abuse-detection and investigation console for a subscription app — detection rules through to evidence, an explainable score and a human decision.
A read-only Spotify library auditor for duplicates, re-releases, suspiciously similar tracks, stale playlists and ghost tracks.
A hair-care platform for iOS built around personalised profiles, curated ingredient data and outputs that explain themselves.
Automated trading systems built rules-first, with hard loss limits, kill switches and simulation-before-live.
A Next.js and Supabase operations dashboard that puts projects, revenue, finances and tasks into one operational view.