Pixel-faithful, from mockup to production.

Figma and Canva mockups rebuilt as real, maintainable code - design tokens, component systems, and ratio-level fidelity. Plus the QA tooling to prove the build matches the design.

Figma Canva Design Tokens shadcn/ui Visual QA

What I do here

01

Faithful reproduction

Designs rebuilt at ratio-level accuracy - real fonts, real spacing, real proportions - not a generic approximation pulled from defaults.

02

Design systems

Tokenized color, type, and spacing wired into component libraries (shadcn/ui, Tailwind) so the system scales past the first screen.

03

Theme reproduction

Full Shopify theme reproductions and brand-palette implementations built section by section from a design spec.

04

Visual QA

Pixel-level visual-regression tooling: multi-run comparison, overlay and heatmap views, and a generated report dashboard.

Recent builds

Brand hero implementation

Client

A Vite, React, and shadcn build implementing a full brand palette and design system from mockups - the public-facing hero of an education brand.

Reactshadcn/uiDesign Tokens
Private client build

Full theme reproduction

Build

A pixel-fidelity reproduction of a premium Shopify theme, rebuilt section by section from the live reference.

ShopifyLiquidFidelity
Build sample on request →

Premium wellness theme (Canva to Liquid)

Client

Canva mockups translated to a custom Liquid theme with ratio-matched layout and the brand's real typefaces, not a stand-in.

CanvaLiquidFidelity
Private client build

Pixel-diff QA tool

Build

A visual-regression testbed: it runs a build multiple times, diffs each against the design reference, and generates an overlay and heatmap comparison dashboard.

Visual QAOpenCVCLI
Internal tool

Tools I reach for

Figma Canva Design Tokens shadcn/ui Tailwind Liquid OpenCV Visual Regression

Other areas of work

Have a design that needs to ship?

Hand me the Figma or Canva file. I will give you back production code that matches it, and the proof that it does.