NELA: AN AI-POWERED RAG-BASED HEALTH ADVISORY SYSTEM FOR REAL-TIME DIAGNOSTICS AND PERSONALIZED HEALTHCARE

ABSTRACT

Current healthcare systems struggle with delayed diagnostics and generic treatment plans. This paper introduces NELA, a Retrieval-Augmented Generation (RAG)-based AI system designed for real-time, evidence-driven health advisory services. Our system integrates diverse datasets by combining synthetic data from ChatGPT GenMedGPT-5k and HealthCareMagic.com with real-world patient interactions to enhance diagnostic accuracy and relevance. NELA generates personalized diagnostic insights within seconds. Evaluated against benchmark datasets, NELA demonstrated a 40% reduction in diagnostic latency and 22% improvement in accuracy compared to rule-based systems. The system addresses ethical challenges such as data bias mitigation through multi-source retrieval and transparent reasoning. Future work explores integration with IoT wearables for continuous health monitoring. NELA bridges cutting-edge AI with clinical workflows, offering a scalable foundation for personalized, real-time healthcare.

Keywords: AI in healthcare, Retrieval-Augmented Generation, real-time diagnostics, clinical decision support, ethical AI