I'm a web analyst. My work is about the website and the people who use it. I use customer data to find where the site can be more accessible and more secure, and turn that into decisions.
Alongside that, I build with AI, in evenings and weekends.
Here are the things I make: an AI assistant that runs my task list, automations that run day and night on my own server, and a knowledge base that makes more than three years of AI conversations searchable.
What I build has to be two things at once: it has to actually run, and I have to be able to explain it. I work at a large public-sector organization, and there you quickly learn that "it works" is only half the story. The other half is being able to explain why you trust it and where the limits are.
That habit is in everything on this site: I don't let AI do the math where a calculator belongs, and nothing goes live without me seeing it first. What that looks like in practice: How I keep my AI in check.
That's why every project here shows what it delivers and what went wrong.
