Used structured ingredient data and strict prompt rules to build an AI skincare assistant, tested on 50 real-world queries for accuracy and logic.

How I am using ChatGPT to prototype UX, prompt flows, and schema logic before training our own model.

I’m leading UX content strategy and system design, while Keon architects the GenAI backend, merging structured data pipelines with LLM outputs.

Skincare Ingredients are complex. We’re building a science-first conversational AI that turns ingredient lists into clear, trustworthy explanations without fear-mongering