I measure prediction markets from the order book up.

I'm finishing a BSc in Software Engineering in Madeira, but most of what I know came from building products that had to work outside a classroom.

I've worked in real estate, creator marketing, esports and market data. That mix taught me to care about more than the code. I want to know who is opening the screen, what they need to decide and whether the product earns its cost.

I build end to end, from the first conversation to the database, API, interface and release, and these days I do it with AI coding agents. The rules they work under run as tests, so their code has to clear the same bar as mine. I also question my own work, and some of my best results were audits that caught a number before anyone relied on it.

That habit is now most of what I do. The research desk I run writes its hypothesis down before it looks at the data, keeps a complete dated record of every measurement, and only promotes a result that survives a confidence interval clustered the right way.