Prismic Page Audit Free online audit tool
Home › About & Contact

About & Contact

I make content-driven websites fast, accessible and searchable — performance, accessibility and technical SEO for Prismic / SvelteKit / Next.js sites.

Start here

Fastest way to start: run the free audit on your site, then send me the result link (or a screenshot).

Run the audit → How to share results

Or browse all tools or notes first.

Who I am

Front-end engineer working remotely for a New Zealand company. My day job covers both building (turning designs into maintainable sites, modelling content for a CMS) and optimising (performance, accessibility, technical SEO).

Stack: Next.js, SvelteKit, Astro, TypeScript, Tailwind CSS. Content: Prismic, Shopify (headless), WordPress. Deploy: Vercel, Cloudflare.

I don't ship templates — structure, content models and interactions are built to spec, and technical SEO plus a speed baseline are part of launch, not an afterthought.

Timezone UTC+8, 4–5 hours of overlap with New Zealand; working with Chinese brands going global and overseas clients, in Chinese and English.

Real numbers, not adjectives

  • This site (prismicaudit.com): desktop Lighthouse 100/100/100/100; HTML first payload 27.7 KB; sitemap 266 URLs; TLS certificate 44 days remaining. Source: local direct Lighthouse (2026-09-30), not a public-tunnel measurement; public access can vary with network latency. Sitemap / TLS read live at render time.
  • Audit sample (client unnamed): Lighthouse snapshot of a New Zealand retail site on 2026-09-18 — LCP 5.1s · TBT 2.04s · 总字节 2090 KiB · Perf 44 / A11y 88 / BP 77 / SEO 83. This is a snapshot / starting point from one audit run — not a claim that I optimized it to those numbers.

See also: home audit · tool directory · notes

Selected work you can actually open

All three are live. Open them and check for yourself instead of taking my adjectives on trust.

Pet Field Guide — a Shopify theme for pet stores

  • 44 sections, 20 page templates, 3 header layouts, 4 style presets, no apps required
  • Tested at six viewport widths: zero horizontal overflow; touch targets at least 24x24 (WCAG 2.5.8)
  • The demo store is open — the password is on the landing page → see the theme

SliceLab — industry site templates

  • Restaurant, clinic, law and general templates on one shared section system
  • Next.js + Prismic (headless): a new industry is a content model plus sections, not a rebuild
  • This is the base I reuse on client builds → see the templates

prismicaudit.com — the tool site itself

  • Server-rendered, two languages, 132 URLs in the sitemap, HTML first payload under 30KB
  • Lighthouse 100/100/100/100 on a detector page
  • The multilingual canonical bug was found and fixed here → read the note

How I help

Website optimization

  • Core Web Vitals
  • Image & font pipeline
  • Third-party script control
  • Technical SEO
  • Accessibility (WCAG AA)
  • Caching & CDN

Website development

  • Custom builds
  • CMS modelling
  • Multilingual sites
  • Headless commerce
  • Animation & interaction
  • Launch-time SEO

What things cost

There is no rate card with hidden tiers. Three fixed prices, and you know what each one costs before you pay anything or wait for a quote.

1. Single fix - $199

  • One specific problem: a slow LCP, a wrong canonical or hreflang, a section that will not drag, horizontal overflow on mobile
  • You get the fixed code and the before/after numbers

2. Audit + fixes - $399

  • A full pass: performance, accessibility, technical SEO, images, structure
  • You get a prioritised issue list plus the first three items actually implemented

3. Monthly care - $79/mo (3 slots)

  • Uptime monitoring, a monthly performance and SEO brief
  • Broken links and certificate checks, plus 2 hours of small changes a month

Not sure which one you need: run the free audit on the home page, send me the result, and I will tell you which tier it is.

How we would work together

  • ① You run the audit (or send me data you already have)
  • ② I return a prioritized issue list + effort estimate
  • ③ After you confirm scope and quote, I start — delivered in batches, each with before/after numbers

What I don't take on

  • Projects that require a guaranteed #1 search ranking (rankings can't be promised; measurable technical metrics can)
  • Full-site rebuilds with no code or admin access — screenshots only
  • Emergency full-site rewrites due in 48 hours

AI-augmented workflow

I run my development workflow on a stack of AI tools, with a clear division of labour:

  • ① Scoping — Claude / ChatGPT turn vague requirements into a task list and acceptance criteria
  • ② Implementation — Cursor in agent mode handles scaffolding, components, refactors and bulk edits
  • ③ Quality gates — automated checks (Lighthouse, accessibility, structured data, link integrity) plus a line-by-line diff review
  • ④ Acceptance & launch — I make the final call: nothing ships until the performance and SEO baselines pass, and monitoring continues after launch

I don't hand off whatever the AI produces — I review, re-test and stand behind every change. That's how one person ships custom, high-quality pages at small-team speed.

Selected work

Screenshot of the Adhesion website home page

Custom build

New Zealand company website · SvelteKit + Prismic

Built end-to-end to spec — structure, CMS content models and interactions. Not a template.

  • SEO 100/100
  • Accessibility 98/100
  • Mobile LCP 2.2s
  • 1,188 KB page weight (all images WebP + responsive)
  • Bilingual, structured data and semantic markup done at launch
Screenshot of the NOMAD LAB storefront

Site build · Shopify + Next.js headless

NOMAD LAB (personal · Shopify + Next.js headless + Tailwind)

  • Shopify as the headless product and cart backend
  • Next.js storefront
  • Dark industrial visual design
Overview page of the Prismic audit tool ({len(sd.DETECTORS)} detectors)

Tools

Prismicaudit (this site)

113 stack detectors + real-browser speed tests; desktop Lighthouse 100/100/100/100

  • Desktop Lighthouse 100/100/100/100 (all four — verify on the home page)
  • 113 stack detectors with hit evidence

Swipe to see more →

How I work

  1. Free diagnosisYou get an audit report — issue list plus impact and fix for each item.
  2. Set targetsWe agree metrics first (e.g. LCP < 2.5s, performance ≥ 90), then I start.
  3. Step-by-step changesChanges on a separate branch so you can roll back; visual changes need your OK first.
  4. HandoverSame measurement before and after to show what changed.

Commitment

Diagnosis is free. If the targets we agreed on in writing aren't met, you don't pay.

Contact

Fill in the form below — no mail app needed. Email and WeChat work too.

Works without opening a mail app.

Open my mail app (recommended)

To：[email protected]　|　Cc：[email protected]

WeChat QR
WeChat: scan to add me
Works from any country. Send a screenshot of the audit and we can talk fixes.