No-Code · Competitive Design · Partner Tools · Revi OS

Inherited five layouts.
Shipped one editor.
Dozens of partners switched to keep using it.

The first version was vibe-coded on Lovable by our CEO and a sales lead. Over five inconsistent layouts, blocked screens, features buried behind features — editing was so hard that even our own team got stuck. I audited every page, benchmarked Shopify, Squarespace, Wix and the restaurant-specific competitor, and rebuilt it into one simple editor restaurant owners actually wanted to use.

Lead Designer · UX Audit · Rebuild
Shopify · Squarespace · Wix · restaurant incumbents
Web · Partner Dashboard · Revi OS
Revi
revi.os / builder
Website builder — media & background panel
Section-level toggle
Each block flips on/off independently. Default layouts still work with everything off. Reveal on toggle
Flip Button(s) on and the full button editor appears below — never before it's needed.
Website builder — typography panel
Use Default · ON Default = one good opinion
Typography inherits the brand defaults. 95% of restaurant owners never touch it — and the site still ships looking polished.
Website builder — font picker open
Use Default · OFF Flip it off, depth unlocks
Full font picker, title + body size, weight. The customization is always one toggle away — never in your face.
Website builder — button styling panel
Deepest customization
Once they're in, operators get the full set — background + text color, subtext, link-to targets — all scoped to this one button, in this one section.

One canvas. One editor. Toggle-gated depth — simple by default, customizable when asked.

The Brief

Five layouts, one editor. Ship the version a restaurant owner can actually use.

The first pass had been vibe-coded on Lovable by our CEO and a sales lead before I was handed the project. They moved fast — but what shipped was five-plus different page layouts that didn't share patterns, screens that blocked users from editing core content, and features stacked on features until even our internal team couldn't demo the product without a cheat sheet.

The brief was simple, the work wasn't: keep the ambition (full website and app creation, multi-brand support, deep customization) but give restaurant owners a single, predictable editor. Something closer to Shopify and Squarespace than to the restaurant-specific incumbent — and crucially, usable by operators who'd never touched a design tool in their life.

What I Inherited

The diagnostic: five layouts, zero consistency.

5+

page layouts

Each new section was a new template. Typography, spacing, and section controls didn't carry between them.

∅

blocked screens

Core flows — hero, menu, team, gallery — had states where users literally couldn't edit the content they saw on the canvas.

∞

nested features

Features hidden three clicks deep inside other features. Powerful if you already knew where to look. Nobody did.

🙃

internal demo-ability

Our own onboarding team needed a cheat sheet to run a live demo. That was the tell.

The Approach

Audit everything. Benchmark the best. Rebuild around one editor.

I started by walking every page in the original build as a restaurant owner would: logging every friction point, every duplicate control, every screen that blocked editing. Then I benchmarked against the tools operators already compared us to — Shopify, Squarespace, Wix, and the restaurant-specific competitor — and catalogued the patterns they'd trained users to expect: left-panel controls, a live canvas, section-scoped editing, smart defaults.

The redesign collapsed five layouts into one. One editor shell, one control panel pattern, one set of section types. Typography, background, buttons, media — each became a tab on the same left panel instead of a separate flow. Deep customization stayed for the operators who wanted it; smart defaults made sure nobody was forced to touch it.

Benchmarked against Shopify Squarespace Wix Restaurant incumbent

What I Shipped

One editor. Deep customization. Multi-brand ready.

01

Collapsed five layouts into one editor shell

Every section — hero, content, media, team, menu, contact — now lives in the same left-panel editor with a live canvas to the right. Users learn the pattern once and apply it everywhere.

02

Unblocked the screens that locked users out

Every editable surface in the canvas now has a one-tap path to the control that edits it. No more "the feature exists but you can't reach it" states. This alone moved the needle in partner feedback.

03

Deep customization with smart defaults

Typography, backgrounds, button styles, media layouts (side-by-side, vertical, grid, scatter), bullet-point colors — real control for operators who want it. Smart defaults generate a working branded site from a name, category, and logo for operators who don't.

04

Multi-brand architecture + website + app in one flow

The incumbent only supported single-brand setups and treated mobile apps as a separate product. I designed one dashboard that handles restaurant groups, franchises, and multi-concept operators — with website and native app as two outputs of the same build.

The Editor

One editor. Four panels. Every customization a restaurant owner actually needs.

Media & background panel — toggle media on/off, choose display style, add buttons
Media & background Section-level toggles. Four display styles. Buttons stay hidden until the operator flips them on.
Typography panel with Use Default toggle on — smart brand defaults
Typography — default on Smart brand defaults mean 95% of owners never touch type, and the site still ships looking polished.
Font picker open — Use Default off reveals full typography customization
Typography — custom unlocked Flip one toggle and the full font picker, title & body sizing, and weight controls appear.
Button styling panel — background and text color, subtext, link target
Button editor The deepest level — background + text color, subtext, link targets — scoped to one button in one section.

Outcome

Dozens switched. And kept building — long after onboarding ended.

Dozens
partners migrated from the incumbent to the rebuilt Revi OS builder
Post-onboarding
partners kept logging back in to build more custom sites for their own businesses
5 → 1
inconsistent layouts collapsed into one predictable editor shell
2-in-1
website + native app created from a single, unified flow

Reflection

The real win wasn't the migration — it was partners coming back on their own.

The thing I'm proudest of here isn't the switch numbers. It's that dozens of operators kept opening the builder after their onboarding call was over, and used it to spin up more sites for sister concepts, pop-ups, and catering brands. That's the test of a simplified tool: do people choose to use it when no one's holding their hand?

Inheriting a product that was vibe-coded and ambitious taught me something I carry into every project now: the AI can write the first version. The designer has to be the one who decides what stays, what dies, and what the user never should have had to think about in the first place.

← Revi OS All work →