This project explores AI as a brand experience layer rather than a standalone tool. I developed an AI skincare assistant concept inspired by CeraVe, designed to guide users through personalized skincare recommendations while maintaining strict brand tone and safety boundaries. This project was independently developed for portfolio purposes and is not affiliated with or officially associated with CeraVe.

The assistant was built using n8n, combined with a Retrieval-Augmented Generation (RAG) architecture and a vector database to ensure responses remain accurate, controlled, and aligned with verified product information. Instead of relying on generic model outputs, the system retrieves structured brand-relevant knowledge before generating responses reducing risk and improving reliability.

From a brand perspective, the assistant was designed to ask structured diagnostic questions (skin type, concerns, environment), educate users without offering medical advice, and recommend simple, barrier-focused routines. Clear conversational guardrails were implemented to define what the assistant can and cannot say, ensuring compliance, safety, and tone consistency.

This project demonstrates my ability to design AI systems that balance technical architecture (RAG, workflow automation, vector retrieval) with brand communication strategy. The goal was not to showcase AI as a gimmick, but to use AI to reduce decision fatigue, simplify product discovery, and strengthen brand trust through structured conversational design.