AutoPercept AI: Pioneering Intelligent Agents in Automotive Design and Engineering
| Reference number | |
| Coordinator | Intended Future AB |
| Funding from Vinnova | SEK 199 995 |
| Project duration | November 2024 - February 2025 |
| Status | Ongoing |
| Venture | Accelerate Swedish partnership |
Purpose and goal
Intended Future is developing intelligent agents for the car industry. In 4 months, we’ll create an MVP of the Perceived Quality (PQ) AI Agent—app to assess vehicle design, reducing iterations by 30%, costs by 15%, and boosting customer satisfaction by 20%. Deliverables: customer journey, backend architecture, prompt engineering docs, workshop summaries, and a software demonstrator. This aligns design with customer expectations, enhancing efficiency, reducing costs, and promoting sustainability.
Expected effects and result
A comprehensive document will outline the customer journey and software experience. The backend architecture will connect the vector database to a stable Large Language Model (LLM), including the full codebase, system architecture diagrams, and technical documentation. It will also feature detailed prompt engineering documentation and a Perceived Quality (PQ) AI Agent demonstrator with limited functionality to showcase its potential to clients.
Planned approach and implementation
The foundational design phase ensures early completion of the customer journey and app experience, making technical development user-centric. The development of the technical backbone includes rapid backend architecture and LLM integration. AI performance validation focuses on accuracy for reliability. Interface readiness finalizes front-end development for interaction and feedback. The project concludes with delivering the MVP demonstrator to showcase the product and gather insights.