Our business is complex, and you don’t become an automotive expert in just a few hours—but EURODECISION quickly grasped our needs. Our project was a bit unclear: we knew we needed rules, but we lacked expertise in algorithm design. It was reassuring to work with artificial intelligence specialists. The EURODECISION consultants are thorough and efficient, and they developed a fully functional algorithm in a very short time.Alisée BRUNOExperimental Psychology Researcher - TROUVOTTO
Founded by an automotive expert and an experimental psychology researcher, the TROUVOTTO startup was born from the observation that no tool existed to help drivers with a purchase project quickly find vehicle models matching their expectations while also considering their psychological needs. On top of this, the tedious task of checking whether dealerships had the desired model in stock—and then visiting to test drive it—only added to the frustration.
To address this gap, TROUVOTTO aimed to provide consumers with an intelligent, brand-neutral questionnaire that would build a personalized profile based on their answers, matching them with suitable vehicle models. This innovative concept won over Bpifrance and French Tech, whose funding enabled the startup to launch its project.
The TROUVOTTO team had the industry expertise to design the questionnaire and related services but lacked the technical skills to develop the car compatibility algorithm for their website. They sought a decision support systems specialist to guide them—EURODECISION’s expertise in artificial intelligence and decision mathematics, their ability to grasp project requirements, and their proven experience with startups quickly convinced TROUVOTTO to entrust them with the project, which began in September 2021.
Initial scoping workshops enabled EURODECISION to precisely gather TROUVOTTO’s functional needs and determine the most suitable technology. The team opted for a rule-based system (specifically the open-source BRMS tool Drools)—an AI solution that models reasoning through logical rules, ideal for decision-making processes based on questionnaires. EURODECISION also proposed an Agile management approach, allowing TROUVOTTO experts to validate progress after each development sprint. Engine testing led to refinements in profile selection criteria, including weighted coefficients and penalty adjustments.
Delivered in December 2021, an initial prototype of the questionnaire was tested by 100 beta users during street marketing campaigns. When asked about their functional (seating capacity, mileage, engine power, etc.), aesthetic (design), and psychological (driving style, etc.) preferences, 75% found the recommendation engine’s suggestions relevant—validating TROUVOTTO’s vision.
EURODECISION then collaborated with ASTRONAUT AGENCY, the team behind TROUVOTTO’s website and user questionnaire interface, to integrate the online engine. The startup’s goal is for website users to connect with a TROUVOTTO advisor, who will refine their selection via phone (e.g., vehicle options). Advisors will then help users locate preferred models at partner dealerships and arrange test drives for two nearby vehicles.
The official launch of the Otto algorithm, backed by a national communication plan, is scheduled for September 2022. Meanwhile, a second testing phase involving nearly 500 participants aims for an over 80% “very high satisfaction” rate. Feedback on the algorithm’s 8.25 trillion possible combinations will further refine the questionnaire, though it already meets TROUVOTTO’s expectations.
