I've been meaning to rebuild my portfolio for a while. Longer than I'd like to admit, honestly. The old version had done its job, but it wasn't telling the right story anymore — and as someone who spent six years at a company doing work I was proud of, I wanted a site that could actually hold all of it.
The Starting Point
Fluid Engine and I Got Off to a Rocky Start
I've been using Squarespace since version 6.0. Before that I was trained on 5.0 support — which, if you've ever seen 5.0, tells you I have a high tolerance for platform quirks. 6.0 and 7.0 were genuinely good. The block system was logical, layouts held across breakpoints, and the whole thing had a calm predictability that let you focus on the design rather than the tool.
Fluid Engine did not have that energy.
The free-form grid sounds appealing right up until you're looking at your tablet view and nothing is where you put it. Elements drift. Spacing that looks intentional on desktop becomes arbitrary in the middle breakpoints. I spent some time trying to make it work the Fluid Engine way. I dragged things around. I nudged. I previewed and adjusted and previewed again. And the results were fine.
Fine is not what I was going for.
The Turn
The Moment I Reached for Code
I don't remember exactly when the switch flipped, but I remember the feeling. I was looking at a layout that should have been simple — a hero section, some project cards, a clean typographic hierarchy — and it looked like a Squarespace site. Not a bad one, but an obvious one. The kind where you can feel the template underneath.
So I opened a code block and started writing HTML.
It sounds like a small thing. It wasn't. The moment I had a div I controlled, with classes I named, with CSS I wrote from scratch, the site started becoming what I had in my head. Not all at once — these things never happen all at once — but the direction became clear.
The Build
Building the System
The first thing I did was define the tokens. Colors, type, spacing — all of it in CSS variables at the top of a global injection block. Simple decisions made once and referenced everywhere. If you want to get into the technical specifics of how that works, I wrote a companion post: Squarespace Doesn't Have to Look Like Squarespace.
/* Brand tokens defined once, used everywhere */ :root { --red: #CB2028; --bg: #FAF7F2; --bg-warm: #F2EDE4; --ink: #1C1714; --ink-mid: #6B5B52; --border: #E4DDD5; --font-serif: 'Aleo', Georgia, serif; --font-sans: 'Inter', Helvetica, sans-serif; }
That single step changed how the whole build felt. Instead of making the same color choice twelve times across twelve pages and hoping it stayed consistent, I made it once. Everything else just pointed back to it.
From there the components came together gradually. A hero class. A campaign block for pairing images with text. A mini case study format for smaller projects that didn't need a full write-up. Each one built once, dropped in wherever it was needed, behaving the same way every time.
It started feeling less like building a website and more like building a design system that happened to live inside Squarespace.
Screenshot — Design System Components
Image placeholder
The Honest Part
The Surprises
A few things were harder than I expected. The blog CSS was a whole thing — Squarespace's own variables kept overriding custom styles in ways that required some creative specificity. Tablet required more deliberate attention than I'd budgeted for. And working through the case studies for my Vernier work — six years of projects I care about — carried more emotional weight than I anticipated.
What was easier than expected was how well the code layer and the platform layer coexisted. Squarespace handles the hosting, the navigation, the blog engine, the form submissions. I handle everything that goes inside the content area. That division of responsibility turned out to be a comfortable one.
Current State
Where It Stands
The site is still in progress — that's the honest answer. Case studies are still being written, images are still being gathered, and there are projects in the queue that haven't been scaffolded yet. A portfolio for a working designer is never really finished; it just reflects wherever you are at the moment.
But it looks like mine. The typography is considered. The color is deliberate. The layouts hold on desktop, tablet, and mobile because I wrote them to. When I share it, I don't feel like I'm handing someone a Squarespace URL — I feel like I'm handing them something I made.
That took longer than expected and more code than anticipated. It was worth it.
Screenshot — melewdesign.com
Image placeholder
The Meta Layer
On Designing for Yourself
There's a particular kind of difficulty in designing for yourself that nobody warns you about. You know too much about what you want, you have no client to push back on your instincts, and every decision is both a design choice and a statement about who you are professionally.
The portfolio is still the best brief I've ever worked on. It has no budget, no deadline, and infinite scope — which means the only thing keeping it moving is discipline and genuine investment in the outcome.
I'm still finding both. But the site is getting there.
