Built with AI. Finished by hand.
A design and accessibility review for products built with Lovable, v0, Claude, Figma Make, and Cursor. AI gets you most of the way there. I help with the last stretch, the part that makes it feel like someone built it, not something.
Free, and no obligation. I'll start with what's already working and we can go from there.
Your build works. Pages load, forms submit, nothing is broken. That's further than most people get.
But there's a stretch between working and convincing, and AI tools stop right at the edge of it. Not because they're bad at design. Because a tool only makes the decisions you ask it to make, and the details that signal someone cared are the ones nobody thinks to ask for.
You care about this thing. That's why you built it. That care just hasn't fully landed on the page yet, and that's true of nearly every AI-built product I look at, including good ones.
What I've learned to look for
-
Designing for everything that isn't the happy path
Most of my work has been on platforms where real people hit real edges, at volumes where every unusual case eventually happens to someone. That teaches you to look at a product and immediately see the paths that were never designed. It's usually the largest gap between a demo and something people can actually rely on.
-
Making it sound like you
There's a difference between copy that's correct and copy only your company could have written. I've spent years translating complicated things into language people can act on, working with subject matter experts to find the specific claim underneath the general one. I'll show you where your product is describing a category instead of itself.
-
Deciding what matters most
A page where everything gets equal weight is a page where nobody was willing to choose. Part of my work has always been sitting between business goals and the interface, figuring out what a product actually needs to communicate first. That translation is a judgment call, and it's the one AI can't make for you because it doesn't know what your business is betting on.
-
Knowing what "finished" looks like
I've built and maintained design systems used across large product surfaces, where the difference between good and shippable lives in details most people never consciously notice. That's the lens I bring here. Not a list of flaws, but a sense of what's still unresolved and which of it a visitor will feel.
-
Accessibility, seriously
I've built WCAG-compliant systems for healthcare interfaces and shipped under state accessibility mandates, where compliance wasn't optional and someone checked. A lot of people building with AI miss these key details. It's the one area where the exposure isn't just aesthetic.
-
Designing for content that changes
Placeholder content is always uniform. Real content never is. Coming from design systems work, I look at a layout and see how it behaves six months from now, with a name twice as long and a list three times the size. It's the difference between something that looks good today and something that stays good.
Who is this for?
-
This is a good fit if
Your product is live, or close to it. There's real content in it, not placeholders. You're the one who'll make the changes, and you want to know exactly which ones are worth making.
-
Probably not yet if
You're still pre-launch with dummy copy, since there isn't much for me to look at.
Send me your URL
Send it over and I'll write back with a few things I noticed. No obligation, and nothing to sit through.
I'll start with what's already working. Most of these end up shorter than people brace for.
If you want to go further after the free look, we'll figure out scope and pricing together. Some products need an hour, some need a week, and I'd rather match the work than sell you a package.