Treffer: Towards a Body of Knowledge for Intelligent Low-Code/No-Code Systems Empowered by GenAI.

The growing complexity of intelligent low-code systems backed by large-scale language models necessitates novel architectural strategies to promote modularity, scalability, and knowledge reuse. This paper proposes a conceptual architecture that separates agents from a structured body of knowledge composed of validated components, enabling their reuse across different contexts. A key contribution of this work lies in the systematic preparation of this body of knowledge, built from artefacts and components extracted from real use cases and tested in production environments. Through this process, reusable elements such as buttons, tables, and interactive charts were identified, validated, and catalogued, ensuring that the body of knowledge is firmly grounded in practical evidence rather than theoretical assumptions. The results demonstrate that this approach not only improves modularity and maintainability, but also establishes the foundations for the continuous evolution of the body of knowledge, supporting advanced integrations such as retrieval-augmented generation and semantic knowledge bases. This highlights the dual role of the architecture: as a structural proposal and as a cumulative process of knowledge consolidation. [ABSTRACT FROM AUTHOR]