AI transformation, in four weeks. Proven every day after.
We send an engineer into your team, ship one AI job into the product, process or system you already run, and leave. What stays is a record your compliance team, your auditor and your board can check any day: what the AI was allowed to do, what it did, and who signed it off.
For regulated businesses. Financial services, healthcare, legal, public sector.
The request
Show us that the Aurora Banking app meets EN 301 549.
From a procurement team · 2 Sep
What you show them
Account overview
42 checks
Signed off
QA lead · 4 Sep
Payments
38 checks, 2 fixes re-checked
Signed off
Product owner · 6 Sep
Onboarding
29 checks
Signed off
QA lead · 9 Sep
Backed by
Built on Google Cloud
Put AI to work. And make sure it's looked after.
Your board wants AI in the product this year, not a pilot that stays a pilot.
Your compliance team wants to know who is accountable for what it does, before it does it.
Your team is already at capacity, and the business case has to travel upward on your name.
Most vendors answer the first line. We answer all three.
In. Build. Leave.
Four weeks, one job, in production. See how a deployment runs, week by week.
In. Week one.
We agree the job, what the AI is allowed to do in it, and who signs. Written down before anything runs.
Build. Weeks two to four.
Our engineer works in your codebase next to your developers. The old way and the new way run side by side, so you can see the difference before you rely on it.
Leave. Week four.
The job runs in production and your team runs it. Every decision the AI makes is still checked against what you allowed and still recorded, on AUDITSU, on a monthly subscription. We maintain the job if you want us to.
What you keep
- The job, live, in your product.
- The code, assigned to you in writing.
- The record of every decision, replayable any day.
- The check, still running.
Most tools show you what's wrong. AUDITSU lets you prove you're right.
Your AI can propose anything. Only the people you name can make it happen. That is what a deployment leaves running: every decision checked against your rules before it takes effect, the check recorded, and a history you can replay any day to the same answer.
Every decision says who allowed it, on what evidence, and when.
It never goes stale. The check runs on every decision, not once a year.
If anyone changes it later, it shows.
In the accessibility product today, nothing a scanner finds reaches your record until a person on your team accepts it. How the record works, and what a deployment leaves running for your job.
The request
Show us that the Aurora Banking app meets EN 301 549.
From a procurement team · 2 Sep
What you show them
Account overview
42 checks
Signed off
QA lead · 4 Sep
Payments
38 checks, 2 fixes re-checked
Signed off
Product owner · 6 Sep
Onboarding
29 checks
Signed off
QA lead · 9 Sep
Who's looking after this?
You deal with Simon and Jason from the first call. One engineer sits inside your team, full time, for the four weeks.
Simon Milner, Founding Architect
Simon designs the system. Twenty-five years in Silicon Valley building the categories others followed: employee number 30 at MMC Networks, which created the network processor and was acquired for $4.5 billion; nine years as Vice President and General Manager at Marvell Semiconductor; two companies founded and sold. He holds a PhD in digital signal processing, and he is the architect of the record behind a deployment.
Jason Crispin, Founder
Jason leads the company and the customer relationship: what the job is, what it is worth, and that it lands. Multiple startup exits, then twenty years winning and delivering contracts worth £12.3 million a year inside a billion-dollar company. He is on your first call and every one after.
Meet the teamOne record. Every job.
We built an accessibility product and found the thing underneath it was the record: who was allowed to do what, what they did, and proof nobody changed it after. Accessibility is live today and on general sale. The same record carries every job we deploy, in every sector we work in.
- Everything we build in your codebase is yours, in writing.
- We work in your repositories with access you can switch off any day.
- Your code is never used to train a model.
Bring one job.
Thirty minutes with Simon and Jason. You bring the job you've been asked to hand to AI, or the thing you've been asked to prove. You leave knowing what it would take, what you'd keep, and what it costs. If the honest answer is not to use AI for it, we'll say so.
Book a callFour weeks. In production. Proven every day after.
What does it cost?
What does it cost?
One job, at a fixed scope, for four weeks, priced on the first call once we have seen the job. You leave that call with a number and the list of what you keep. Auditing a website is free. Apps and managed audits are priced on the call, from £1,500 for a managed audit.
How much of our team's time does it take?
How much of our team's time does it take?
Three people: a named owner for the job, a developer who can review, and the person who will sign. Week one is the time it takes to agree the job and the rules in writing. After that it is the review your developers already do, and a standup.
What if it does not work?
What if it does not work?
The old way keeps running beside the new one for the whole build, so you can stop at any point. You keep the code, the rules and the record of what was tried. And if the honest answer on the first call is not to use AI for this job, we say so and there is no deployment.
Where do our code and data go, and who can see them?
Where do our code and data go, and who can see them?
We work in your repositories, on your identity provider, with access you can revoke any day. Your code never trains a model. If the job cannot go to a hosted model at all, we run it on hardware you own or a UK cloud you control.
Who owns the code and the rules afterwards?
Who owns the code and the rules afterwards?
You do. Everything built in your codebase is assigned to you in writing, and the rules for what the AI may do are yours to change. The check and the record run on AUDITSU, on a monthly subscription, so they keep running after we leave.
How is this different from a dev shop, or from giving the team an AI coding assistant?
How is this different from a dev shop, or from giving the team an AI coding assistant?
A dev shop leaves you code. An assistant makes one developer faster. Neither leaves you able to answer who allowed what the AI did. We ship one job into production and leave the check running: what the AI may do is written down first, every decision is checked against it, a named person signs, and you can replay the history any day.
What does “proven every day after” mean?
What does “proven every day after” mean?
Not a case study. It means the check does not stop when we leave. Every decision the AI makes is still checked against your rules before it takes effect and still recorded, and if anyone changes the rules or the record later, it shows. Patent pending, UK application GB2620101.2.
What happens after the four weeks?
What happens after the four weeks?
Your team runs the job. The check keeps running on AUDITSU and the record keeps growing, on a monthly subscription. When you have a second job, the record underneath is the same one.