RAG ENGINE INFORMATION RETRIEVAL SYSTEM USING LLM

RAG ENGINE INFORMATION RETRIEVAL SYSTEM USING LLM

Mr. S. DHINAHARAN

Jagadheesan S, Maheshkumaar K, Harivel K, Yugapathy R K

BACHELOR OF TECHNOLOGY – 4th YEAR

DEPARTMENT OF ARITIFICIAL INTELLIGENCE AND DATA SCIENCE

SRI SHAKTHI OF ENGINEERING AND TECHNOLOGY(AUTONOMOUS)

COIMBATORE-641062

ABSTRACT

The RAG Engine Information Retrieval System using Large Language Models (LLM) is designed to improve the accuracy and relevance of information retrieval by combining retrieval-based methods with generative AI models. Traditional AI models generate responses based only on pre-trained knowledge, which may lead to outdated or incorrect answers. To overcome this limitation, the proposed system integrates a Retrieval-Augmented Generation (RAG) approach. The system retrieves relevant documents from a knowledge base using vector embeddings and similarity search techniques. These retrieved documents are then provided as context to a Large Language Model, which generates accurate and context-aware responses. The system is developed using modern technologies such as Python, vector databases, and LLM APIs. The RAG engine enhances response accuracy, reduces hallucination, and ensures up-to-date information retrieval. It is scalable, efficient, and suitable for applications such as chatbots, question-answering systems, and knowledge management platforms. In addition, the system is capable of handling complex user queries by understanding semantic meaning rather than relying on simple keyword matching. This improves the overall quality of search results and ensures that users receive more relevant and meaningful information. The integration of embedding models enables the system to capture contextual relationships between words, making the retrieval process more intelligent and efficient.

Keywords: RAG, Information Retrieval, Large Language Models (LLM), Vector Database, Semantic Search, Text Embeddings, Retrieval-Augmented Generation, Natural Language Processing (NLP), AI Chatbot, Context-Aware Response, Knowledge Base, Similarity Search, Generative AI, Data Retrieval, Intelligent Systems, Document Retrieval, Query Processing, Prompt Engineering, Transformer Models, Deep Learning, AI-based Search Systems, Context Injection, Retrieval Pipeline, Hybrid Search, Dense Retrieval, Indexing Techniques, FAISS, Pinecone, Scalable AI Systems