ai transformation

AI systems that
run your growth work

We build the systems with your team: research, content, outreach, and reporting, checked before anything ships. Your team owns them and keeps them running after we step back.

Scoped to your team · three-month minimum

what you get

Five parts,
one system

01 · Context

One shared brief

Your positioning, customers, voice, and rules in one place. Every system reads it, so everything sounds like you.

02 · Growth skills

One for each growth function

From ad research to outreach. Each one turns how your team works into steps AI can follow.

03 · Agents

Agents with quality checks

Research, content, and reviewer agents. Every output is scored against a clear standard before a person approves it.

04 · Connections

Built into your tools

Connected to the tools you already use, from your CRM to Figma and Slack.

05 · Team training

Your team runs it

Hands-on workshops and weekly office hours, so your team can run and extend the systems.

See AI training →
how it works

Context, build,
check, hand over

Most AI projects stall at the first draft. Ours run in four stages, and each one leaves something your team keeps.

01 · Context

Teach the AI your business first

Generic AI writes generic work. So we start by writing down what makes yours different: who you sell to, what they care about, who you compete with, and how you sound.

  • Interviews with your growth leads
  • Your best-performing ads, pages, and emails as reference
  • One shared brief every system reads before it writes
brief.mdExample

# Positioning
Payroll for UK hospitality groups with 5–50 sites.

# Customers
Ops directors. Hate rota chaos and late pay runs.

# Competitors
Generic HR suites, spreadsheets.

# Voice
Plain, warm, specific. Numbers over adjectives.

# Never
"Revolutionary". Exclamation marks. Fake urgency.

02 · Build

Turn your playbook into skills

A skill is a written process AI can follow: when to run, which steps, which tools, and the rules it must keep. We write one for each growth job, starting from skills that already run in our own work, then connect them to your stack.

  • Built from processes we already use in our own work
  • Skills for ad creative, alternative pages, content, and reporting
  • Connected to your CRM, design files, analytics, and Slack
See the tools we build with →
skills/ad-creativeExample

Runs when "make this week's Meta ad variations"

Reads brief.md, your top ads, competitor ads from the ad library

Steps find the angles that are winning → write 10 hooks → render image ads in your brand → send to review

Rules one claim per ad · brand palette only · real product shots

Claude Code Figma Apify Gemini Linear
03 · Check

AI that reviews AI

Speed is easy with AI. Quality is the hard part. Every system gets a reviewer agent that scores each output against a written rubric. Anything under the ship bar goes back for another draft, automatically, before a person sees it.

  • A scoring rubric for every output type
  • A ship bar and hard gates that can't be argued with
  • A person approves what matters, in Slack
reviewer · "vs competitor" pageExample
Names why people switch9
Fair to the competitor8
Sounds like the brief9
Proof behind every claim6
Ship bar 8/10Redo: back the pricing claim
04 · Hand over

Yours to run and extend

We don't leave a black box. Every skill is written down, every system has an owner, and your team learns to run and change it before we step back.

  • Every skill and agent documented in plain English
  • An owner on your team for each system
  • Training and office hours while your team takes over
handover checklistExample

✓ Brief and skills in your workspace

✓ Reviewer rubrics and ship bars set

✓ Weekly report posting to Slack

✓ Owner named for each system

✓ Team trained to add new skills

what changes

Before and after

Before After

Positioning scattered across old docs

One shared brief every system reads

Never enough content

A content pipeline with checks built in

AI tools bought, barely used

AI systems your team uses every day

One keen person's workflow nobody else can run

Systems the whole team runs and extends

Agency work that stops with the retainer

Systems you keep after we step back

is it for you?

You're a good fit if

A good fit
  • You want AI systems your team owns and extends
  • You need more output without adding headcount
  • You want quality standards on everything AI produces
  • Your team has had training, or doesn't need it, and the work still runs by hand
Why three months

It takes that long to build the systems, train the team, and get real buy-in. Shorter engagements end before the systems are part of how the team works.

see the starting point

Our growth skills
are free

Positioning, customer research, prospecting, outreach, and ads. Every engagement starts from them.

questions
How is this different from an agency?

An agency runs the work for you, and it stops when the retainer stops. We build it with your team, and you keep every system.

How is it different from AI training?

Training teaches your team to build. Transformation builds the systems with them, end to end, with quality checks and a handover.

What do we own at the end?

Everything: the shared brief, the skills, the agents, and the documentation, in your own accounts.

Do you need access to our accounts?

Yes, to connect the systems to your tools. Everything runs in your accounts, and you own it.

How do you handle our data?

Your data stays in your tools. We use only the access each system needs, and we never use one client's work in another's.

Which tools do you build on?

The ones you already use: Claude or ChatGPT, your CRM, Slack, Figma, Airtable, and the rest of your stack.

How much of our team's time does it take?

An owner for each system, and time for the weekly office hours. We do the heavy lifting in between.

What does the first month look like?

We build the shared brief first, then the first system. Months two and three add the rest and hand them over.

How do you check quality?

Every system gets a scoring rubric and a ship bar. Output that misses the bar is redone, and a person approves what matters before it ships.

What happens if something breaks after handover?

Every system is documented, and your team is trained to fix and extend it. We check in 30 days after handover.

Will this replace people on our team?

It takes on the repeat work, so your team spends its time on the work that needs them.

Why three months?

It takes that long to build the systems, train the team, and get real buy-in. Shorter engagements end before the systems are part of how the team works.

What if our team hasn't had AI training?

Training is part of the engagement. Teams that are far behind start with AI training instead.

How much does it cost?

A monthly fee, scoped to the systems you need, with a three-month minimum. We send a fixed quote after the growth audit or a short call.

Can we start smaller?

Yes. Start with the growth audit, or with the AI training programme.

let's build something

Get your team's
week back

Tell us about your team, your tools, and the growth work that still runs by hand. We'll come back with a scope.

Talk to us