Etudiant : TAMMAL MOHAMED

Filière : Master Smart Industry (M2SI)

Encadrant : Pr. NAJI MOHAMED

Annèe : 2025

Résumé : This internship project focuses on building a smart hybrid Retrieval-Augmented Generation (RAG) system that combines the power of Large Language Models (LLMs), graph databases, and vector search to provide more accurate and context-aware answers to user questions. The system automatically classifies user questions and selects the best retrieval mode, vector-based, graph-based, or mix to both, based on the type of the question. During this work, I explored key concepts such as RAG architecture, knowledge graphs, dense and sparse retrieval, and fine-tuning language models like DistilBERT for query classification. I also gained hands-on experience using tools like Hugging Face, Streamlit, and others. This project helped me strengthen my skills in natural language processing, machine learning, and system evaluation using tools like RAGAS. Overall the internship provided valuable technical and practical experience that shaped both my understanding and capabilities in AI-powered systems.