A MULTIMODAL CONVERSATIONAL AI SYSTEM FOR INTELLIGENT REAL-TIME ANDROID INTERACTION

Abstract— Artificial Intelligence (AI) has significantly transformed Human-Computer Interaction (HCI) by enabling natural language-based communication between humans and machines. Voice assistants are helpful tools that make our daily tasks easier and less complicated when we use digital systems. However, most existing assistants suffer from limited personalization, lack of contextual understanding, platform dependency, and poor real-time conversational flow. This project introduces a conversational AI voice agent designed to understand and respond to users through speech in a real-time manner across both desktop and Android mobile platforms. Keywords—HCI, Large language model (LLM), API, TTS, Conversational AI, LiveKit, Gemini, Voice Assistant, Speech Interaction, NLP, Android, Realtime Streaming.