IntraNote: Prototyping a Design Rationale System with AI-driven Reflective Reasoning

We demonstrate IntraNote, short for InteRaction Annotator, a reflection tool that explores new possibilities for Design Rationale Systems (DRS). Collaborative design plays a crucial role in architecture design process, where multiple designers discuss and interact with 2D or 3D design artifacts physically or virtually. As this discussion and the context of interaction fade from memory over time, existing 3D CAD workflows lack support to capture and revisit these design actions. Current DRS can document structured decisions and textual summaries but typically neglect semantic content and real-time interactions, especially in 3D CAD environments. Recent advancements in Large Language Models can track and interpret iterations in dialog, presenting a chance to bridge this gap. To explore this opportunity, IntraNote synchronously records dialogues and 3D object interactions, and generates timeline-based interaction tracking, LLM-driven semantic annotations, and customized causal inference. It demonstrates how generative AI can support reflection, understanding, and recall in collaborative design.