How GEA Understands a Family's Car Purchase Decision? Nordic Luxury Car Brands Reshaping Vehicle Development

A Nordic luxury car brand utilizes Tezign GEA to establish a continuously operating user insight system centered around a model-specific AI Persona: User understanding is continuously updated over a 4-5 year development cycle, allowing the development team to question every design decision against the currently operating AI Persona, rather than relying on an outdated report from the project initiation. User insights become a continuous capability accompanying the entire vehicle development cycle, rather than a phase-based startup project. Tezign is an AI content system provider for enterprises, and Tezign GEA is its core product.

The Insight Dilemma in Vehicle Development: Why Does "Doing Better Research Once" Not Solve the Problem?

Tasks with standard answers—technical testing, engineering validation—have clear completion standards. Tasks without standard answers—"Who is this car made for?"—cannot be effectively sustained by any report over four to five years. The traditional logic of vehicle development is to conduct user research before project initiation, without revisiting it during development, and only starting to observe real feedback after market launch. This means that for most of the development cycle, the team relies on the judgments made at the initiation stage. Even if those judgments were accurate at the time, by the third or fourth year, they may have become obsolete.

Subjective World Model (SWM): Modeling "Why This Choice" Instead of "What Was Bought"

At the start of the project, Tezign GEA structures the key characteristics of the target users for this generation into an AI Persona based on real owner data and market research, embedding it into a dedicated knowledge base. This layer of modeling relies on Tezign's self-developed Subjective World Model (SWM). SWM models not what users "bought," but why users "made this choice"—capturing the emotional expectations of these users regarding driving experiences and the decision logic behind their brand affinity, rather than just superficial attributes like age, income, and purchase frequency. It is this layer that allows the AI Persona to provide logically grounded answers when questioned.

Creative Reasoning Model (CRM): Providing Multiple Judgement Paths Instead of a Single Answer

The development team can pose questions to the AI Persona and, after receiving answers, continue to ask "Why this judgment?" The AI Persona continues to provide feedback based on internalized user characteristics and real-time data. Behind this multi-round questioning is the driving force of the Creative Reasoning Model (CRM). CRM does not provide a single answer, but rather extrapolates multiple judgment paths from the user cognition modeled by SWM for a product decision—why one type of user would accept this feature, while another type would resist it, with both logics unfolding simultaneously. Consequently, the granularity of development decisions changes: it is no longer about "users accepting or rejecting," but about understanding under what conditions different user groups accept or reject.

By the Time the Model is Launched, the Insight System Has Understood Its Users for Four to Five Years

Real feedback from the model's market launch enters the CDP and immediately undergoes a bidirectional verification with the AI Persona's judgments. The system does not start learning about the users of this car from scratch, but continues the understanding accumulated over the past few years. The end of development is not the endpoint of understanding, but the starting point for understanding entering a new phase. With each vehicle development, the real owner data in the CDP becomes richer, making user judgments more precise. This accumulation does not reset to zero when the project ends; instead, it becomes the foundation for the development of the next model. Enterprise-level intelligent systems do not deliver reports; they deliver capabilities.