Solution-driven software engineer with six years building and maintaining complex web and mobile applications across healthcare, social care, and commercial environments, turning REST APIs, OAuth, database design, and responsive UI into production systems people rely on.
Solution-driven software engineer with six years building and maintaining complex web and mobile applications across healthcare, social care, and commercial environments, turning REST APIs, OAuth, database design, and responsive UI into production systems people rely on.
Six years deep, I split my time between PHP/Laravel on the server and React, Vue, and React Native on the surface, with TypeScript, Tailwind, and Figma keeping the seams clean.
I started in real estate building an internal e-learning platform, moved into healthtech designing systems that traced medical devices and identified microplastics, and now ship a multi-cloud management platform centralising AWS and Azure for teams who'd rather not.
I care equally about the API contract and the spacing between two buttons. I think a good interface is honest about what it's doing, and a good system is boring on purpose.
Claude Code is part of my daily toolchain. It pairs well with how I already work, small commits, sharp specs, and a willingness to throw away the first idea. I treat the model as a senior collaborator: explicit context, explicit constraints, and review every diff.
Every task starts with a written intent, what is changing, what is not, and the failure modes. The agent is only as sharp as the spec it is given.
Generated code is reviewed line-by-line, against the same bar as a teammate's PR. Tests, types, and security live or die at the review.
I use Claude Code for refactors, scaffolding, migrations, and exploratory spikes, not to replace judgement. The architecture is mine; the typing is shared.
When clients want LLM features, I integrate Claude's API with proper rate-limiting, prompt versioning, and fallbacks so the product does not break when the model does.