The questions we hear at the first meeting.

Why is your audit paid?

A free audit has to pay for itself through the mission that follows, so its conclusion always leans the same way. Ours is paid to decide, which leaves it free to conclude NO-GO.

You keep the four audit documents whatever happens: its assessment, the costed roadmap, the specifications and the framing note for your executive committee. The one-hour diagnostic with our founder remains free: it is the conversation that precedes the audit.

How much does an audit cost?

From €4,000 to €20,000 before tax, depending on scope: how many processes are examined, how many sites are involved, the state of the data. That amount covers the diagnostic, the audit and the GO / NO-GO decision, and it is fixed before launch.

The build budget is only discussed afterwards.

What happens if the audit concludes NO-GO?

The project stops there and no build budget is committed. Three situations lead to that conclusion: the necessary data does not exist or cannot be used, the target process changes too often to be automated, or the expected gain does not cover the cost of running it in production.

You leave with the conditions to meet before reopening the subject later.

How long before a first system is running?

The audit takes weeks, the first production release takes months. The exact timeline depends on the state of your data and on how many people need to be involved on your side, two things we measure before committing to anything.

The roadmap is ordered so that a first project can start without waiting for a decision on the rest.

How are you different from an IT services firm?

An IT services firm sells man-days and bills for time spent. We sell a monthly retainer fixed at signature, on a written scope with written deliverables.

If the work takes more effort than planned, that is on us. Your CFO knows the cost of the mission the moment it is signed.

And compared with freelancers?

With us, the audit, the development, the production release and the follow-up are shared between several specialists, who document as they go. If one of us stops, the mission continues and you have no one to replace.

Why not hire in-house?

Hiring is a lasting commitment, after a search, an onboarding and a ramp-up. The retainer starts by the month and stops at the agreed term.

The two combine well: we build and document, your teams take over the code once the system is in service.

How is a mission billed?

As a monthly retainer, from €9,000 before tax. It covers design, development, the production release, documentation and follow-up once the system is in service.

The expense books as an operating cost, with no capitalization and no borrowing, and your debt capacity stays available for your other investments. Ailix Consulting holds the French CII accreditation: the work you entrust to us can qualify for the innovation tax credit, within the limit of eligible expenses.

Who owns the code?

You do. The source code, the deployment scripts and the documentation are handed over, and another team can take them up. Nothing we build runs on a platform we would keep the keys to, and no per-user license is added after go-live.

Are your systems GDPR-compliant?

Compliance rests with you, the data controller. Our job is to give you the means to demonstrate it. The infrastructures we design for our clients are hosted in France, at Scaleway, and the distinction matters: hosting is one thing, calls to the models are another. On sensitive processing, calls to the models are configured with Zero Data Retention: the provider keeps no record of the requests.

We provide the traceability elements your DPO needs for the record of processing activities. In healthcare, the architecture is installed with an HDS-certified host, France’s certification for health data hosting: Ailix Consulting is not a host.

Does our data go to a foreign provider?

That depends on the model selected, and it is a decision made during the audit. When nothing may leave France, the architecture is designed for that from the start.

When a foreign model is selected, the call is configured with Zero Data Retention and the exact scope of the data transmitted is written into the framing note, before go-live.

What happens when the AI gets it wrong?

It gets things wrong, and the system is built so that this shows. Answers cite their sources, human validations sit at the points decided during the audit, and decisions are logged so they can be replayed.

On anything sent to a client, a patient or a citizen, we require human review. The “Verify” step of the method deals precisely with controlling hallucinations and with real adoption.

Do we have to replace our current tools?

No. The agents and automations plug into the ERP, TMS, CRM or messaging your teams already use.

Replacing an ERP or a CRM is a project in its own right, and we do not mix it into the AI mission.