I wrote my capabilities, my four case studies and my target role before I opened a single builder. AI is fast at execution and terrible at guessing your strategy.

I began in Canva. It gave me a decent deck and no AI website widget — 'coming soon.' Knowing what a tool can't do is worth the afternoon it costs you.

Wix Studio was faster to start, but the template had a point of view my work didn't fit into. If you're bending your story to fit a layout, change the layout.

A former colleague, Tony, built his portfolio with Claude while I built mine with Lovable. Two experiments, one conversation, twice the learning.

+52% adoption. +124% sessions. 60%+ CTR lift. AI can shape a sentence beautifully, but the proof has to be yours. Never let it invent a metric.

I asked for design directions, not 'a website.' Palette constraints, tone, composition, and a refusal — what I did NOT want — got me something usable on the first pass.

Hiring leaders skim. Situation, Action, Results with bolded keywords beats three beautiful paragraphs nobody finishes.

I rewrote every AI draft out loud. If it didn't sound like something I'd actually say in an interview, it didn't ship.

Alt text, meta descriptions, headline variants, a first pass at a messaging framework. Boring work is where AI earns its keep.

I bought a vanity domain before the site existed. A vanity domain is the cheapest signal of seriousness you can buy.

A few of my work samples are still placeholders. Posting the build-in-progress gets better feedback than a perfect launch nobody saw coming.

Positioning, proof, channels, cadence. My portfolio got the same GTM treatment I'd give any product — because I'm the product this quarter.