NewsMind AI: A Retrieval-Augmented Generation System for Intelligent News Analysis

NewsMind AI is a Retrieval-Augmented Generation (RAG) based system designed for intelligent and real-time news analysis. The system integrates GNews API for live data ingestion, Sentence Transformers for semantic embedding generation, and FAISS for efficient vector similarity search. A Groq-powered Llama-3 Large Language Model is used to generate context-aware and accurate responses. The architecture follows a modular pipeline including data preprocessing, embedding generation, vector indexing, and retrieval-based response generation. The system significantly reduces hallucination issues found in standalone LLMs by grounding responses in real-time data. Experimental results demonstrate improved accuracy, relevance, and reduced hallucination rates compared to traditional keyword-based search and standalone LLM approaches. This work highlights the effectiveness of Retrieval-Augmented Generation in building scalable and reliable AI-powered information retrieval systems.