Artificial intelligence FAQ for business

Xperience AI is an independent artificial intelligence consultancy based in Meyrin, canton of Geneva, co-founded by Sébastien Brunelle and Antoine Doury. We support small and mid-sized companies and other organisations across French-speaking Switzerland, France and internationally, throughout their AI transformation: maturity assessment, costed roadmap, team training, governance and project ownership. More than 30 organisations supported and over 500 employees trained to date.

On three points. Independence: we neither sell nor resell any technology or licence. The person in front of you: the senior consultant who sells the engagement is the one who delivers it, with more than twenty-five years of digital and AI transformation experience gained in large corporations. The end result: we deliver recommendations that can be executed with your actual resources, and that work with or without us.

We work across Switzerland, particularly in the French-speaking cantons (Geneva, Vaud, Valais, Fribourg, Neuchâtel, Jura), throughout France and internationally. Training sessions and co-creation workshops are preferably held on site, because team buy-in is won in the room. Assessment, findings presentations and follow-up can be conducted remotely.

Yes, and that is often where our work delivers the most value. Individual use of consumer tools generates real but scattered gains, unmeasured and sometimes non-compliant, a phenomenon known as shadow AI. At that point the issue is no longer access to the technology, but its integration into processes, data governance and the measurement of results.

It is a structured evaluation of your capacity to create value with artificial intelligence, measured across six dimensions: ambition and strategy, organisation and skills, data and governance, technologies and tools, culture and buy-in, existing use cases. You receive an overall score and a score per dimension, a map of your current uses and risks, and a findings meeting with your decision makers. The method is described in full on our AI audit and consulting page.

The lightest format is a digital questionnaire sent to contributors, which takes around fifteen to twenty minutes per respondent. Information can also be gathered through interviews, in person or by video call, of roughly thirty minutes each. We adapt the approach to your teams’ actual availability.

Three levels of commitment, three verbs. AI Essential to get started: one day, fixed fee of CHF 1,900 excl. VAT, for owners of very small businesses and independents with fewer than five staff. AI Impulse to frame the programme: one month, from CHF 6,500 excl. VAT with staged payment, for companies ready to build their AI vision together with their teams. AI Transform to steer it: the framing phase from CHF 6,500 excl. VAT, then ongoing support from CHF 2,500 excl. VAT per month depending on the level of involvement. The full line-by-line comparison is on our AI audit and consulting page.

Both routes work, and many of our clients start with training. Half a day of AI fundamentals before the ideation workshops means the roadmap discussion begins with a shared language, a realistic view of the limits and a better sense of what is possible. If your leadership team is already aligned on the subject, the maturity assessment can be commissioned directly.

It is the Chief Digital and AI Officer function, carried out part time by an external practitioner: chairing the steering committee, arbitrating priorities, holding the roadmap, enforcing the governance framework and measuring adoption. For a small or mid-sized company, hiring that profile full time is rarely justified or even possible. Fractional access provides the capability without the cost or the delay of a recruitment process. This is the content of the ongoing support phase of AI Transform.

Return on investment is not declared at assessment stage, it is built initiative by initiative. For every use case identified, our method systematically quantifies the estimated gain (hours freed up, additional revenue, fewer errors, cost savings) and the effort required (implementation, licences, change management), then calculates a payback horizon. In our clients’ context, our assumptions model a potential programme return of between five and ten times the investment over three years. This is a model built on assumptions discussed with you, not a guarantee of results.

Because the obstacle is organisational before it is technological. The MIT NANDA report, The GenAI Divide: State of AI in Business (2025), observes that 95 % of generative AI pilots produced no measurable impact on the profit and loss account, mainly for want of integration into processes and continuous learning. Our method addresses those causes directly: initiatives selected on a quantified gain and effort pairing, training for the teams concerned, change management, and adoption measured through indicators defined before launch. No project is started without an identified business owner.

No, and waiting for perfect data is one of the main causes of inertia. Many high value use cases, such as assisted writing, document search, automation of administrative tasks or document analysis, do not depend on a structured data warehouse. The assessment separates the initiatives that can be run immediately with what you already have from those that require a data prerequisite, and places the latter further out on the roadmap.

We sign a confidentiality agreement before starting whenever you wish. Documents shared with us are limited to what is strictly necessary, hosted in environments whose location is disclosed to you, and deleted at the end of the engagement on request. Where the context requires it, the engagement can be conducted with no transfer of real data at all, on the basis of interviews, process descriptions and anonymised documents. The terms are set out in our privacy policy.

In training, we recommend working on your own documents and processes rather than on generic examples. During scoping we agree together which access rights and which documents may be used, within your own security framework, and we systematically recall in session the applicable data protection rules.

Compliance is built in from the framing stage, not bolted on at the end as an option. Every initiative on the roadmap is qualified against the applicable texts: the revised Swiss Federal Act on Data Protection (nFADP), the GDPR as soon as processing concerns individuals located in the European Union, and the European artificial intelligence regulation (Regulation (EU) 2024/1689, the AI Act) for companies placing AI systems on the EU market or using them there. Use cases falling into the high risk category are subject to a dedicated analysis of obligations and mitigation measures before any development begins. These same subjects are covered within each of our training programmes, not relegated to an appendix.

Potentially yes. The AI Act applies on an extraterritorial basis: a company established outside the Union falls within its scope as soon as it places an AI system on the Union market, or where the output produced by that system is used within the Union. A Swiss SME selling into France or handling files for European clients therefore needs to qualify its exposure. We carry out that qualification during the assessment.

AI whose uses are documented and explainable, whose data handling is controlled and compliant, whose impact on jobs is anticipated together with the people concerned, and whose resource consumption is factored into the choice of solutions. In practice this takes the form of an AI charter, written rules of use, a business owner for each use case, and human oversight of any decision that affects a person.

Through a combination of awareness sessions, collaborative workshops and training, guided by one principle: the people who will use the tools take part in identifying the use cases. Adoption is then measured through indicators defined before launch, not observed after the fact. Resistance to change is not inevitable, it is the consequence of a project designed without the people who will live with it.

Thirty minutes, with no commitment and no charge, to understand your context, answer your questions and tell you which package fits best. If none of them fits your need, we say so and point you elsewhere.

Fixed, published and agreed before we start. You know what you are committing to before you sign. Quoted fees exclude taxes and travel costs beyond the canton of Geneva, and engagements in France are subject to specific pricing. For the ongoing support phase of AI Transform, the monthly fee varies with the level of involvement agreed at the outset. If you prefer, we can also work on an hourly basis with a monthly cap set in advance.

Yes. We are continuously building a network of trusted partners in Switzerland and France: developers, integrators, data specialists, sector experts and training providers. We are also looking for subject matter experts who want to co-create a thematic track with us, along the lines of the Skills 2030 track built with Capital Énergie. If you work in one of these fields and our approach speaks to you, get in touch.

Contact us for a free conversation and we will help you to decide. A half-day fundamentals session is also a low-commitment way to test how we work before engaging on anything larger.