Services · + AI & automation
AI, built into how you work
Every business has work a machine could do: orders keyed from one system into another, an inbox sorted by hand, the same numbers pulled into a spreadsheet every Monday. We find where that time goes, then build AI and automation into the tools you already use. Nothing has to be replaced.
Where the time goes
The work nobody should be doing
These are the jobs people bring to us. Each one is a few hours a week somewhere in the business, done by someone who was hired for something else.
- 01
Orders and stock
Orders retyped from the store into the accounts system, stock counted in two places, tracking numbers copied back by hand.
What we do The systems talk to each other, and a job checks every order went through. What can't be matched goes to a person, with the reason.
- 02
The inbox
Customer and supplier emails read, sorted and answered one at a time, by whoever is free.
What we do Each message is read, classified and either answered from your own records or drafted for someone to approve. Your tone, your rules.
- 03
Quotes and invoices
Prices looked up, quotes typed, invoices chased, all by hand.
What we do Quotes drafted from the enquiry and your price list, invoices raised and chased on your schedule, with a person signing off where it matters.
- 04
Reporting
Numbers pulled from three systems into a spreadsheet every week, then explained in a meeting.
What we do The figures land in one place on their own, with the week's changes written up in plain English and the odd ones flagged.
- 05
Product content
Descriptions, specifications, translations and alt text written one product at a time.
What we do Drafted from your data in your voice, checked by you, and kept in step when the product changes.
- 06
Customer support
The same twenty questions, at every hour, with the answer somewhere in the order history.
What we do An agent that reads the message, looks the customer up and resolves what it can in your voice, and hands the rest to a person with the context attached.
How it works
Map it first, then build
The same four steps as our Systems work, because that's what this is: the systems a business runs on, with AI doing the parts that used to need a person.
- Map
A week with you, looking at how work actually flows: what's keyed in twice, what waits on a person, what a machine could do instead.
- Propose
A short list of what to automate, what each saves and what it costs, before anything is built. Some of it won't need AI at all, and we'll say so.
- Build
Into the tools you use already, tested against real orders and real messages, with a person checking the results until it has earned its trust.
- Run
Monitored, with a log of every decision it makes, and changed as the business changes.
Already running
Built here, in use today
Travel eSIM · our own venture
CocoRoam
A WhatsApp support agent trained on real support history, and order jobs that find and fix failed purchases before anyone notices.
View case study ↗AI SEO platform · our own venture
AISEO
Reads Search Console, analytics and the store, ranks the fixes by what moves revenue, writes the content and creates and runs the ads. Built for our own work, now a product.
Visit ↗
Sports distribution · B2B
Specialist Sports
Stock, orders and customers kept in step with Business Central in real time, so nothing is keyed in twice.
View case study ↗
Bikes · made to measure
Atherton Bikes
A custom ERP that tracks every component of a made-to-measure frame from order to dispatch.
View case study ↗Questions we get asked
Before you ask
- Do we have to replace our systems?
- No. We build into what you already run, whether that's Shopify, Magento, Business Central, Xero or a shared inbox. Replacing a system is a last resort, and we'd tell you if we thought you needed to.
- What does the first step involve, and how long does it take?
- About a week. We sit with the people doing the work, follow an order, an enquiry and a week's reporting from start to finish, and write up where the time goes. You get a short list of what to automate, what each saves and what it would cost. You can take that list anywhere.
- Is this a chatbot?
- Usually not. Most of what saves time happens behind the scenes: matching, drafting, checking, moving data between systems. A customer-facing agent is one of the things we build, when it's the right one, and it's connected to your real records rather than a script.
- Which AI models do you use, and where does our data go?
- We pick the model for the job, and we tell you which. Your data stays in your systems; the model sees only what a task needs, under a business agreement that doesn't train on it. We'll put the data flow in writing before anything is built.
- What happens when it gets something wrong?
- Everything it does is logged, anything with money or a customer behind it goes past a person first until it has earned trust, and a wrong answer is one you can trace, correct and stop from happening again.
- Where do most businesses see the first saving?
- Wherever the same information is typed in twice. Orders and stock between the store and the accounts, and the inbox, are the usual places. Both are usually the quickest to pay back, and both are where we'd start.
- How big a company does this suit?
- It suits a business with a few people doing repetitive work they'd rather not, which is most of them. The map comes first, so a small business isn't sold more than it needs.
Next
Where does your week go?
Tell us in a few lines and we'll say what we'd look at first: [email protected]
Put a brief together →