Rebuilding my website with AI as a collaborator

For a long time, my website never really felt like mine. Like many designers, I started with WordPress and Semplice when it first appeared on the market. It wasn’t perfect, but it gave me everything I needed at the time: a place to showcase my work, iterate on my portfolio, and slowly build my personal brand without having to think too much about the underlying technology.

As my career evolved, though, that setup started to feel increasingly disconnected from the way I actually work. I spend my days designing systems that are scalable, intentional, and built to evolve over time, yet my own website relied on themes, plugins, and architectural decisions that I didn’t fully understand or control. I could change how it looked, but I never really owned the foundations. Eventually, that gap became too obvious to ignore.

Rather than continuing to patch an ageing setup, I decided to rebuild the website from scratch. Today it’s built around GitHub, Vercel, and a headless CMS, giving me complete ownership of every part of the experience. Every component, design token, interaction, and content model exists because I designed it that way, not because a theme expected it to. It feels much closer to the environments I’m used to working in professionally, where everything is versioned, structured, and designed to evolve rather than being constrained by hidden layers of complexity.

What surprised me most, however, wasn’t the technology itself but the way I approached building it. This became the first project where AI genuinely felt like a collaborator rather than simply another tool. I worked almost exclusively with ChatGPT and Codex throughout the rebuild, using them to explore implementation approaches, prototype components, refine interactions, review code, and challenge my own thinking. They didn’t replace the creative work or make design decisions for me, but they dramatically reduced the friction between having an idea and seeing it come to life.

That shift changed the rhythm of the entire project. Instead of spending hours setting up boilerplate or getting blocked by implementation details, I could move quickly into refinement and experimentation. The boundary between design and development became much less rigid, allowing ideas to move naturally from concept to working component. Rather than replacing creativity, AI gave me more time to focus on it by removing much of the overhead that normally slows personal projects down.

Looking back, this rebuild wasn’t really about leaving WordPress behind or adopting a more modern technology stack. It was about creating a workspace that genuinely reflects how I think about products today. The same principles I apply when building design systems for organisations—intentional architecture, scalability, maintainability, and continuous evolution—are now embedded in my own website. For the first time, my portfolio doesn’t feel like something I maintain; it feels like a product I can continue building and improving for years to come.