Services

Software with AI in it for organisations of every size, and the governance that keeps that AI accountable. The same small team designs it, builds it, puts it live and stays with it afterwards.

We build

Software with AI in it, from a website for a small business to the platform a whole operation runs on, and every AI feature we ship keeps a record of what it did.

Two developers at a desk in the evening looking at a dashboard on a large monitor.

Platforms and SaaS products

Products with accounts, roles, billing and an AI feature people actually use.

We build the whole thing: the data model, the sign-in, the payments, the admin side your staff run it from, and the AI feature that gives it a reason to exist. It's built to run for years and to change without breaking.

Every model call the product makes is logged and checked, so when a customer or an auditor asks why the system did something, there's an answer on file.

Includes

Sign-in, accounts and roles
Payments and subscriptions
An admin side for your staff
AI features with every call logged

A village pub landlord checking a table booking on his phone behind the bar.

Business websites and apps

Sites and apps for small and growing businesses, with the parts that bring in work wired in.

Bookings, payments, enquiries and reminders, quick on a phone and easy to keep up to date yourself. Built so search engines can find every page, and so it still works when the person who set it up has moved on.

Where it helps, an assistant answers common questions from your own information and passes anything it can't handle to a person.

Includes

Online booking and payments
Enquiry forms that reach the right inbox
Pages search engines can find
Content you can update yourself

An operations manager at her desk in a London office at dusk, reviewing an AI assistant's chat and its list of completed tasks on a large monitor.

AI assistants and agents

Assistants that answer from your own documents, and agents that carry out steps for your staff.

Each one works inside limits you set: what it can read, what it can change, and which actions wait for a person to approve them. Every step it takes is recorded, so you can see what it did and why.

Where the data can't leave the building, we run the model on your own hardware.

Includes

Answers drawn from your own documents
Agents that carry out tasks in several steps
Approval steps for actions that need a person
A record of every step taken

Two colleagues in a warehouse office studying an orders dashboard on a large monitor, shelves of boxes behind the glass.

Internal tools and dashboards

The spreadsheet the business runs on, turned into a proper tool.

We start from the spreadsheet or the process as it runs today, keep what works, and build the tool around the people who use it, so nobody has to relearn their job.

Sign-in, permissions, an audit trail of who changed what, and reports that used to take an afternoon to put together.

Includes

Sign-in and permissions
A record of who changed what
Reports on demand
Connections to the systems you already use

We govern

AI governance for organisations using AI in decisions that matter: a record of what each system did and why, checks before it acts, and reporting that holds up in front of a board or a regulator.

Three senior colleagues reviewing a bound report at a walnut boardroom table, the City of London at dusk behind them.

AI governance reviews

Where AI is in use across the organisation, what each system decides, and who answers for it.

We find the AI you know about and the AI you don't, from the model behind a customer chat to the scoring tool a supplier switched on. Then we write a short report on what to fix first, in the order the risk calls for.

The review covers AI you've built, AI you've bought, and AI you're about to switch on.

Includes

An inventory of the AI in use
Who is accountable for each system
Checks on suppliers and models
A report on what to fix first

A printed AI activity audit report on a desk, its decision log showing sealed records with the details blurred.

Audit trails for AI decisions

Every model call, prompt and output kept in a tamper-evident record.

Each record is signed and chained to the one before it, so a decision can be reconstructed months later and nobody can quietly change the history. It's the same approach our own products use.

We can add the trail to systems you already run, or build it into new ones from the start.

Includes

Signed, chained records
Prompts, outputs and model versions kept
Decisions reconstructed on request
Records your own team can check

A compliance manager about to approve a decision on screen while a colleague points to a checklist.

Guardrails and human sign-off

Checks that run before an AI output reaches a customer, and named people who approve the decisions that need a person.

We set out which decisions the AI can make alone, which need a reviewer, and what happens when a check fails, and we build those rules into the system so they hold on a busy day.

Includes

Checks on every output before it goes out
Named reviewers for decisions that need a person
Clear rules for when a check fails
A record of every approval

A printed report on a desk beside a cup of coffee and a pair of glasses.

Board and regulator reporting

Plain-English reports on how your AI is governed, written for the people who have to answer for it.

What the systems do, what was checked, what went wrong and what changed as a result, drawn from the records and not from memory.

Includes

Reports for boards and committees
Evidence packs for regulators
Incident reviews
Guidance and training for staff

Sectors

The engineering is the same wherever it lands. What changes is what the sector asks of it, and we build to that.

Ways to work with us

Four shapes of engagement. Most clients start with one and move to another as the work grows.

Fixed-scope build

A defined website, app or platform, priced once it's scoped and delivered in stages you can see working.

Ongoing development

A monthly arrangement for teams that need new features, fixes and support week to week.

Governance review

A fixed-length review of the AI you use, ending in a written report and a plan in order of priority.

Advisory days

Senior time by the day, for a decision, a workshop with your team, or a second opinion.

How we work

No long discovery phase before anything gets built.

A plain conversation

You tell us what you need, and we tell you what it would take, including when the honest answer is a smaller build.

A working first version

You see real software early, on a private link, and shape it from there.

Governed as it's built

Every AI feature logs what it did and stays inside the limits you set.

Live and looked after

We put it live, watch it, and keep it current after launch.

Where to start

Two ways in, depending on how far along you are.

Intelligence Score

Five questions that give your organisation a score out of 100 for AI maturity, with its strengths and gaps.

Intelligence Brief

A written brief on where AI could help, and what it would take to govern it.

A good fit

You want working software early, not a long document first
You'd like the same people to design, build and look after it
You use AI in decisions about customers, money or staff
Your board or a regulator has started asking questions
You want evidence, not a policy that sits in a drawer
You'd rather have a straight answer on cost and time

Questions

How much does a build cost?

It depends on what it needs to do: a business website is a different job from a platform. We give you a fixed figure once it's scoped, and the quickest way to get there is to send a project brief.

How long does it take?

A business website usually takes weeks and a platform takes months. The project brief gives you an estimate before anything is agreed.

Can you work with the systems we already use?

Yes. Most builds connect to something already in place, such as a booking system, a payment provider or a CRM.

What happens after launch?

We can look after it: hosting, updates, fixes and improvements, under an ongoing arrangement if you want one.

We only use off-the-shelf AI tools. Do we still need AI governance?

Yes, if those tools touch decisions about customers, money or staff. AI you buy is still your responsibility.

What do we get at the end of a governance review?

A written report: what AI is in use, the risks, who is accountable for each system, and what to fix first, in order.

Does it cover the rules that apply to us?

We map your AI against the rules that apply to your sector and tell you where you stand. We aren't a law firm, so we work alongside your legal advisers.

Start a project

Tell us what you want to build, or which AI you need to govern, and we'll come back with how we'd approach it.

Websites, apps and platforms
AI assistants and agents
AI governance reviews and audit trails
A UK registered company

Send us a project brief, or take the five-question Intelligence Score first.

Start a project Take the Intelligence Score