DermaGlow AI
AI skincare advisor that turns a short skin questionnaire into a personalized routine and product picks.
- Next.js
- TypeScript
- AI API
- Tailwind CSS
The problem
Skincare advice online is either generic listicles or product marketing - neither accounts for the one thing that matters: your actual skin. Real personalization needs to reason over skin type, concerns and goals together, which is exactly the kind of unstructured-input problem AI models are good at.
What I built
DermaGlow AI runs a three-step flow: users fill out a short form describing their skin type, concerns and objectives; an AI model processes that into an individual skin profile; and the app returns a tailored skincare routine alongside concrete product suggestions.
The whole app is built with Next.js and TypeScript, with a clean Tailwind CSS interface that keeps the form-to-routine journey friction-free.
Architecture
A Next.js App Router application: the questionnaire submits to a server-side route that prompts an AI API with the structured profile and shapes the response into routine steps and product recommendations the UI can render consistently.
Results
- A short form is all it takes - the AI handles the reasoning from profile to routine.
- Recommendations come back structured (routine steps + products), not as a wall of generated text.
- End-to-end TypeScript keeps the AI response shape honest between server and UI.
Stack
- App
- Next.js · TypeScript · Tailwind CSS
- AI
- AI API integration