This project focuses on developing a Multimodal Retrieval-Augmented Generation (RAG) system for personalized recommendations in the fashion and cosmetics domains. The system integrates advanced AI techniques, including LLMs (Large Language Models) and multimodal embeddings, to suggest eco-friendly and trending products based on user preferences. By combining textual descriptions, product images, sustainability certifications, and trend data, the system creates a comprehensive, user-driven recommendation platform.
Key components of the project include a Flask-based backend for API management, a React frontend for user interaction, and a vector database to store multimodal embeddings efficiently. Leveraging technologies like OpenAI API, Gemini, and embedding services such as Hugging Face, the system enables real-time recommendations with enhanced accuracy and relevance. This platform aims to bridge the gap in sustainable fashion and cosmetics by offering environmentally conscious, trend-aware, and user-personalized suggestions through a user-friendly web interface.
| Name Surname | Student Number | Github Account |
|---|---|---|
| Hazal KANTAR | 202111036 | hazalkntr |
| Ahmet Doğukan GÜNDEMİR | 202111033 | dogukangundemir |
| Ali Boran BEKTAŞ | 202111001 | Boranbektas |
| Hikmet Berkin BULUT | 202111057 | bekX0 |
- Dr. Serdar Arslan