ARCHITECTURE × AIGC · OPEN TO COLLABORATION
I work between architecture and AI, turning contextual knowledge, spatial analysis, design generation, and compliance review into usable creative workflows.
Explore open-source work01 / OPEN-SOURCE WORK
A multi-agent workflow for historic urban districts, connecting spatial typology analysis, RAG-based rule retrieval, design generation, and compliance review through an integrated interface for knowledge, versioning, and reporting.
View on GitHubA tailored design platform that brings site research and concept generation into one workflow, exploring how generative AI can support design within a specific cultural and regional context.
View on GitHubPORTFOLIO / VISUAL FRAGMENTS
Anonymous architectural imagery, selected as visual notes on material, context, and form.
02 / APPROACH
My current public work centers on the UAD AIGCDesign series: from a V1.0 prototype that connects site research with design generation to a V2.0 system built around RAG, multi-agent collaboration, and iterative compliance review. I want AI to do more than produce an image—it should participate in design judgment, knowledge retrieval, and the full cycle of exploration.
Structure spatial types, design rules, and cultural context as knowledge that a model can retrieve and reference.
Connect analysis, generation, risk annotation, and iteration in a transparent, traceable loop.
Start from the designer’s process so AI remains understandable, controllable, and genuinely useful.
03 / IN PROGRESS
Developing integrated tools that move from site input and knowledge retrieval to design output.
Exploring more reliable orchestration and state transfer across analysis, generation, and review roles.
Turning design research into products that can run, communicate an idea, and continue to evolve.
04 / CONTACT
Whether it concerns AIGC design tools, architectural prototypes, or a remote coffee, I'd be glad to hear from you.