ReadAssist: An AI-Powered Intelligent Reading and Interactive Insight Generation System

The rapid growth of digital content has increased the demand for intelligent tools that improve reading comprehension and knowledge retention. This paper presents ReadAssist, an AI-powered intelligent reading and insight generation system designed to enhance the digital reading experience through real-time contextual assistance and personalized learning support. The proposed system combines viewport-aware context extraction, hybrid AI inference, deterministic caching, and dynamic user interface generation to provide features such as contextual explanations, multilingual translation, knowledge blocks, quizzes, readability assessment, and smart note generation. A hybrid architecture utilizing local large language models through Ollama and cloud-based Gemini services ensures both privacy and scalability. Experimental evaluation demonstrates significant performance improvements through Redisbased caching and efficient context extraction strategies, enabling responsive and interactive user experiences. By integrating AI-driven educational assistance directly into the reading workflow, ReadAssist transforms passive document consumption into an active and engaging learning process.