You have decided to put an AI deployment into production. The value is clear. Before it goes ahead, the people
accountable need answers to four questions: what could go wrong, what the deployment is required to satisfy, which controls follow, and what is still unresolved.
The AIQURIS Deployment Assessment Sprint produces that in three weeks, for five hours of your people’s time.
Briefing to completed assessment
Your people’s time, across three meetings
Decision-ready assessment
Fixed, excl. tax
You have policies, regulations, standards and tools. None of them says what this specific deployment requires. The business owner believes the risks are manageable. Legal is not sure what applies. Security has questions nobody has written down. Nobody disagrees exactly, and nothing moves.
You nominate one deployment. Everything else is scheduled around a single three-hour block in your calendar, with a short briefing before and a readout after.
A short structured briefing on the deployment you have chosen, with your business owner and one technical contact.
AIQURIS arrives with the assessment already derived from the briefing: the risk profile, the requirements that apply and the controls they call for, on screen. Your people confirm the facts, challenge anything that does not match what they know, and decide what closes the gaps.
The assessment is finalised in the register on the AIQURIS platform, with the controls as agreed in the workshop. The report is generated from that record rather than written up afterwards, and we present the findings and the controls in a one-hour session.
The assessment states what this deployment requires and where it stands against those requirements. That is the basis your risk, legal and security functions have been asking for, in a form they can accept or challenge.
Each control answers a risk this deployment actually presents, and states the evidence that satisfies it. Nothing generic, and nothing to translate from a framework first.
Every unresolved item is recorded with an owner and what closes it. Nothing is buried, and nothing has to be rediscovered at the next review.
Your team assessed this deployment on the AIQURIS platform, with a founder in the room. Platform access runs 12 months, so the next deployments go through the same route without us.
The assessment stays live in the register on the AIQURIS platform, and is issued as a written report for the people who were not in the room.
Fixed, excl. tax. One deployment, from briefing to completed assessment. No scoping exercise, no day rates.
Everything needed to decide whether this deployment proceeds, and on what conditions.
Control implementation, deployment changes and reassessment sit outside the Sprint.
The full S$9,800 is credited against a CONTROL+ engagement started within 90 days of issue. CONTROL+ then assesses your evidence and determines the residual risk. Going on is optional and does not change the Sprint’s conclusions.
Led by Dr Andreas Hauser or Dr Martin Saerbeck, who built TÜV SÜD’s global AI quality practice and sit on the committees writing the standards. Deployments assessed across regulated and safety-relevant sectors, from public transport to pharmaceutical manufacturing.
Four Sprints run each month, led by a founder. Next available dates on request.
Not sure it is the right deployment? Request a 30-minute call.
Disclaimer. The Sprint is an assessment, not a certification or approval. AIQURIS is not an accredited conformity assessment body. It works from what you confirm during the Sprint: it does not test the AI system, and does not verify that controls are implemented or requirements met. Evidence verification and residual risk sit with CONTROL+. Your organisation owns the decision to proceed, and on what conditions.
Three hours, from four people: the deployment owner, accountable for getting it live. Risk or legal, who reviews the assessment or answers for it later. Security or IT, who knows what can go wrong and what the system touches. And someone who can decide, or who can obtain a binding decision. Depending on the deployment, data protection, procurement or the team that built the system may need to be there too. Attendance is confirmed at the briefing, because without those roles the assessment cannot be tested against what each function knows.
The one closest to a decision. A deployment already in production works as well as one waiting to launch, and one that is stuck in review is usually the most useful of all. We will help you pick during the briefing.
No. This assessment works from a description of the deployment: its purpose, users, data, operating context and what the system is permitted to do. No system access is required for it.
Unknowns are recorded as unknowns rather than assumed. They appear in the assessment as open items with an owner, which is usually more useful than a guess.
Yes. The session works remotely provided the right people are present and engaged for the full three hours.
That is the point of running it. The output states what the deployment requires and what is still open. What you do about it is your organisation’s decision.
Not as a Sprint each time, and it is not meant to. Use the first as a calibration deployment, then assess the rest on the platform at the depth each one needs. Discuss a portfolio rollout.
Already familiar with the method? The same assessment is available self-service on the platform, without the briefing, the preparation or the facilitated session. See RISK+ →
That is the one to run first. Tell us what it is and we will take it from there. Or book a 30-minute call →
