15+ years making complex systems work — in capital markets, compliance, and enterprise fintech. The technology changes. The discipline doesn't. Now I'm applying that same rigour to building SaaS products — with AI integrated where it genuinely makes them better.
I spent 15 years in the trenches of financial systems — capital markets platforms, compliance engines, ETL pipelines, wealth management tools. Not talking about them. Building them. Systems that couldn't fail. Code that ran at 3 AM on schedules no one would notice until they didn't run.
I'm not just exploring AI. I'm shipping it — in production apps with real users, real data, and real consequences. The same engineering discipline that kept trading platforms stable is what I bring to every product I build.
That's the difference between shipping something and shipping something that works.
15 years across enterprise and startup environments leaves a wide technical footprint. Here's the honest picture of what I work with.
Emergency ID card platform for travellers, seniors, and families with special needs. QR-coded cards give first responders instant access to critical medical information — 20+ language support, group management, multi-country users. AI is integrated to speed up one specific step: when users upload medical documents, vision models extract structured data with a user-review confirmation before anything is saved. The product stands on its own. AI makes that one step faster and less error-prone.
Invitation-only knowledge and membership platform for sports clubs and professional organisations. Multi-tenant architecture with per-group branding, access control, and content management. Built in direct response to a real customer need and shipped as a working MVP within weeks.
Dynamic QR code tracking and redirect platform with real-time scan analytics, scheduled redirects, password-protected codes, and a public REST API. Features a built-in MCP server — making it natively usable by AI agents via Claude and other MCP-compatible tools. The first QR platform designed for the AI agent ecosystem.
The actual reference documents and component libraries I use when starting a new project. No email required — just download and use them.
38 design decisions every production web app should enforce from day one — with a description of why each one matters and a concrete implementation note.
Download PDFHow to structure AI-assisted builds — from a single prompt to a full multi-session project. Six prompt habits, eight anti-patterns, a copy-paste template, and a staged build methodology.
Download PDFCurated public repositories — tools, starters, and reference projects I've built and opened up. Take what's useful. Raise an issue. Fork it.
38 production web app standards implemented as ready-to-use React components. Tailwind v4, Vite, zero UI library dependencies. Drop it in and have a running demo in under two minutes.
Five developer utilities that run entirely on your local machine — env file comparison, JWT decode and generate, cryptographic key generation, and local directory backup. No data sent anywhere. Built for the workflows where that actually matters.
Following along, have a question, or just want to say hello — I'm happy to hear from you.