AI-
Enabled
Designer

AI has changed how I design and build. I still lead with the problem, the user and the details, but I can now take an idea from first sketch to a launched product on my own.

Here's how AI fits into each stage of my workflow.

This portfolio and the projects on this page were built entirely with AI.

Built with Claude Code

When I moved to Porto, I used Claude Code to build small tools for settling in: learning Portuguese, understanding what a mortgage really costs, and finding a home near a good school.

Web app Mortgage calculator Ten banks, taxes and fees in one dashboard, ranked by what a loan really costs. Open project iOS / Android Memeo The flashcard app I use to practise Portuguese every day. Open project Web app Escolas do Porto A map that scores neighbourhoods around Porto on schools and travel times. Open project

AI-native workflow

AI now runs through every stage of my process. What used to need a team and weeks now takes me days. My current setup looks like this:

Research & Define

Claude or ChatGPT, Notion AI, Fathom, Miro

Market, audience and competitor insights, plus the rules and data behind a product, uncovered in minutes instead of days of digging. User interviews are recorded and summarised automatically, so I can focus on the conversation.

Ideate & Prototype

Lovable, Figma Make, Claude

Explore concepts and alternative directions in minutes, then turn the best ones into working prototypes to validate ideas early.

Design & UX Copy

Claude Design, Figma MCP, Figma Assistant, Nano Banana, ComfyUI (motion)

Screens and UX copy with Claude Design and Figma's AI assistant, on a design system Claude reads directly through the Figma MCP. Imagery comes from Nano Banana, and motion is explored in ComfyUI.

Build & Ship

Claude Code, GitHub Copilot, Xcode

Turn designs into working web apps and native iOS and Android apps, live on the web, the App Store and Google Play, without waiting for engineering.

Measure & Improve

Claude or ChatGPT, Mixpanel, Hotjar

Track and watch how people actually use the product, summarise feedback and reviews, spot what to fix next, and ship improvements in days rather than sprints.