Most tools give you a log. We give you a record you can replay.
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 to the same answer.
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
A log is not proof.
- When the AI has done a thousand things, a log tells you what happened.
- It cannot tell you whether each one was allowed at the time.
- It cannot tell you whether the log itself has been changed.
- The evidence sits in the vendor’s platform, in the vendor’s format.
- Nobody can answer “who allowed that?” without reconstructing it by hand.
Proof is not a better log. It is a record of permission, replayable by anyone you show it to.
Who allowed this, on what evidence, and has anything changed since?
What the AI is allowed to do is written down first.
Every decision is checked against those rules before it happens.
A named person signs it off. The check produces the evidence; a person judges.
You can replay the whole history any day and get the same answer.
If anyone changes it later, it shows. Patent pending, UK application GB2620101.2.
That is what a deployment leaves running for your job, and what the record holds. In the accessibility product today, a person on your team accepts every finding before it reaches your record.
What is live, and what a deployment leaves behind.
We would rather tell you this plainly than have you find it out afterwards.
Live today: accessibility
Guided checks for every page and screen, a ticket for every failure, and a statement built from the results. Nothing a scanner finds reaches your record until a person on your team accepts it, and every verdict is recorded against the person who made it.
In early use: document sign-off
A named reviewer and a named approver on every document, with no self-approval. An approval is bound to the version it was given for, so if the content changes afterwards, it shows.
What a deployment leaves running
The check before each decision, against your rules, and the replay of the whole history. That is what our engineer wires into the job we build with you. Replay is not yet a screen you log into, and it is not something you can switch on inside the accessibility product today.
A statement built this way, live.
Our own accessibility statement is produced from our own audit in AUDITSU, and it names what is still open. It is the nearest thing to a record you can read without booking anything.
Who this is for.
- Compliance and risk leaders who have to answer for what the AI did.
- Engineering leaders who would rather enforce the rules than write them down twice.
- Boards and auditors who want the same answer every time they ask.
Not for
A dashboard of what your AI has been doing. This is the evidence of what it was allowed to do.
Who you deal with.
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
He designed the record: what the AI is allowed to do, checked before it acts, and replayable afterwards. Twenty-five years in Silicon Valley before that.
Jason Crispin, Founder
He owns the customer side of a deployment: what the job is, what it is worth, and that it lands. He is on the first call and every one after.
What people ask.
How is this different from an audit log?
How is this different from an audit log?
A log records what happened, after it happened. A record says what was allowed before it happened, who allowed it, on what evidence, and lets you replay the history to the same answer. A log cannot tell you whether each action was permitted, or whether the log itself has changed.
Who decides what the AI is allowed to do?
Who decides what the AI is allowed to do?
You do, in week one, in your words. The rules are yours, versioned, and changed by your team with a sign-off recorded against each change.
Does the software decide whether we comply?
Does the software decide whether we comply?
No, and it never will. The check produces the evidence; a named person judges. That is true in the accessibility product today and in everything a deployment builds.
Can our auditor or regulator see it?
Can our auditor or regulator see it?
You can replay the history and show the same answer any day. There is no separate auditor login yet, so today that means showing it to them rather than handing them a seat.
What is patented?
What is patented?
The method behind the record: what an AI is allowed to do, checked before it acts, and provable afterwards. The application is pending, so we never say patented.
What do we run, and what runs on AUDITSU?
What do we run, and what runs on AUDITSU?
The job and the code are yours to run: everything we build in your codebase is assigned to you in writing. The check and the record run on AUDITSU, on a monthly subscription, and the rules stay yours to change. We maintain the job if you ask.
Where to go next.
What have you been asked to prove?
Thirty minutes with Simon and Jason. Bring the request, and we will tell you what you could show and what it would take.