Start a project

How we use AI

Almost everyone in our industry uses AI now, whether they say so or not. Here’s exactly where we use it, where people stay in charge, and how it lets you build on your own website.

Our position

We use AI to keep quality up and prices down. Nearly twenty years of experience decides what goes live.

Where we use it

An extra pair of hands, not a replacement for judgement.

Design

AI designs your pages from your brief and brand. You leave comments right on the design, and we refine them with you until they’re right.

Writing and reviewing code

AI drafts and reviews code alongside us. Every change is read, tested and signed off by a person before it goes live.

Testing, constantly

Automated checks for security, accessibility, speed and broken links run on every change, far more often than anyone could by hand.

Documentation that keeps up

Every site comes with notes on how it’s built and the rules it follows, written for people and for AI assistants alike.

Experience first

AI-forward, not AI-only.

AI is fast, but it’s not infallible. Jordan has been designing and building websites since 2008, and knowing what good looks like is what keeps AI’s mistakes off your site and lets us use AI properly for you.

Security

first: checked on every change, by a person

WCAG 2.2

AA, the accessibility standard we build to

Since 2008

building websites, long before AI could help

Build on your site

BYOB: Build your own blocks – with AI.

Need something new after launch? Build it from your current AI, and we’ll check it and put it live.

How BYOB works

No big quotes for small changes

A new section months after launch doesn’t mean waiting on an agency quote. Build it that week.

No new tool to learn

Work in Claude, ChatGPT, Antigravity or whichever AI you already use.

Checked before it goes live

We review every block for security, accessibility, brand and SEO before it reaches your visitors.

$200 a block

One flat price for each block, so you know the cost before you start.

Who we work with

Why we use Claude, by Anthropic.

No AI company is perfect, and we’ll keep watching how the industry behaves. We chose Anthropic for how it behaves as a company, and how seriously it takes safety and ethics, not only for how good its models are.

Safety is its purpose

Anthropic is a public benefit corporation, set up with the safe development of AI as its mission.

It publishes its rules

It publishes the principles its Claude models are trained to follow, and a Responsible Scaling Policy that sets the safeguards required before more capable models are released.

Your data isn’t training data

We don’t put your private data into AI tools, and we use them on settings that keep our work out of model training.

Questions

What people ask us about AI.

Will AI design my website?

Yes, and it’s one of the reasons our websites cost less. AI produces the designs from your brief and brand, guided by us. You review them and leave comments right on the design. We work through every comment with AI until it’s right, and nothing is built until you sign it off.

Is AI-written code safe?

Only as safe as the review behind it. That’s why every change, ours or yours, is checked by a person for security before it’s merged, and tested again once it’s live.

What if I’d rather AI wasn’t used on my project?

Then we’ll leave it out. Jordan has been designing and building websites by hand since 2008, long before AI could help, and still knows how to do it well.

A project built without AI takes longer, so we’ll quote it separately and you can choose what suits you. Either way, you get the same care for security, speed and accessibility.

Want a website you can build on?

Tell us what you’re making, and how you’d like to work with AI. We’ll show you what’s possible.

Start a project