How do you measure the results of an AI rollout?

With four metrics: time saved per employee, real adoption (active users, not accounts created), the quality of the answers produced and consolidated ROI, tracked continuously.

Marianne, Atout France's trip planner that builds personalized itineraries to discover France

For the Paris 2024 Olympic Games, France.fr wanted to inspire visitors to discover French destinations. Marianne builds a tailor-made itinerary from the visitor's preferences, in 18 languages, and checks it against Datatourisme data.

Industry: Travel & hospitality
Roles: Marketing

Key figures

Personalized travel itineraries, available in 18 languages.
Suggestions checked against Datatourisme data before they are offered.
An itinerary visitors can edit, share, export as a PDF or post on a France.fr community wall.

Atout France's need for this project: offer an innovative, personalized travel experience that invites tourists to discover more of France.

The context: Paris 2024 and visitors from around the world

With the Paris 2024 Olympic Games ahead, France.fr wanted to inspire visitors to discover French destinations with an innovative offer.

Offer personalized travel itineraries, available in 18 languages, to meet the needs of international visitors.
Offer a richer, intuitive user experience that makes exploring destinations easier.

The solution: from the visitor's preferences to a tailor-made itinerary

The trip planner starts by understanding the traveler: their interests, their budget, the places they want to see. The AI uses this information to generate a tailor-made itinerary.

Suggestions checked before they are offered

Before an itinerary is shown to the visitor, it is compared with Datatourisme data. This check ensures the recommendations are relevant and accurate.

An itinerary that stays in the visitor's hands

Once the itinerary is generated, visitors can add, replace or remove stops in a simple interface. The finished itinerary can be shared, turned into a PDF or posted on a France.fr community wall, so other visitors can benefit from it.

Agents deployed

What runs in production

Geni-Marketing

Marianne: trip planner

Builds a personalized itinerary from the visitor's preferences, in 18 languages, and checks every suggestion against Datatourisme data.

18 languages

for international visitors

4 months

of development

5 to 10 people

on the GENIAL team

Key takeaways

What this story tells tourism players

An AI-generated itinerary is only worth something if it is accurate. By checking every suggestion against a reference dataset before offering it, Marianne combines the flexibility of personalized recommendations with the reliability of official data.

Do your visitors find the trip that suits them?

Thirty minutes to imagine, on your own tourism data, recommendations that are both personal and reliable.

Book a meeting Free AI self-assessment

Frequently asked questions

Scoping takes 2 to 3 weeks and ends with a prioritized roadmap; deploying the first agents takes 4 to 6 weeks. The first measurable time savings usually show up after two to three months.