Cir4fun

Xgility (XGL) plays a leading role in the Cir4Fun project by designing and implementing an advanced Generative AI-based Decision Support System (DSS) tailored to the needs of sustainable furniture eco-design. Our work primarily focuses on leading the development of a custom Large Language Model (LLM) trained on both public and project-specific datasets.

Our contributions include:

• Development of a Generative AI Framework: We customize and fine-tune open-source LLMs to offer intelligent suggestions for material selection, design optimization, and lifecycle strategies, ensuring sustainability and circularity are built into every decision. We integrate Cir4Fun domain knowledge, eco-scoring systems, and lifecycle assessment outputs into the AI model, enhancing its decision-making capability with structured knowledge and standards-based ontologies.
• Conversational Interface for Designers and Stakeholders: Our AI tools support natural language interaction, allowing designers and manufacturers to explore “what-if” scenarios and get recommendations for circular design and production strategies. We collaborate closely with industry stakeholders to iteratively improve the AI assistant, validate outputs against expert knowledge, and ensure usability in real-world manufacturing contexts.
• Deployment and Training Support: Xgility contributes to deployment strategies, user training, and educational materials to ensure broad adoption of AI solutions across the furniture value chain.
• Stakeholder App Development: Leading the creation of a web and mobile app that enables data capture, lifecycle tracking, interaction with Digital Product Passports (DPP), and access to training materials.

Through our leadership in AI innovation, Xgility helps Cir4Fun bridge the gap between complex data ecosystems and actionable insights—empowering sustainable transformation in the European furniture industry.