Feng Shui AI takes a photo of a room and returns a structured analysis based on classical feng shui principles — energy flow, furniture placement, color balance, and the relationship between space and the bagua map. The goal was to take a domain that typically requires an expensive in-person consultation and make it accessible through a simple upload. Computer vision handles the spatial interpretation: identifying walls, furniture clusters, entrances, and natural light sources before the analysis layer applies feng shui reasoning to what it finds.
Building it required translating a deeply qualitative practice into something a model could evaluate consistently. That meant encoding feng shui principles as structured prompts with grounding rules, rather than asking the model to improvise. The output is opinionated and specific — not generic wellness advice, but room-by-room recommendations with clear rationale. It's a useful case study in applying AI to domains that feel subjective but have enough underlying structure to make consistent analysis possible.