CureCall: AI-Powered Healthcare Chatbot
Atharva Pagar · 2025
With the growing reliance on digital healthcare systems, accessibility remains a significant challenge, especially for individuals with disabilities or limited technological proficiency. Traditional chatbots rely heavily on text-based interfaces, restricting their usability across diverse demographics. This paper introduces CureCall, an AI-powered multimodal healthcare chatbot that allows users to interact using voice and medical images, delivering voice-based, doctor-like responses. CureCall integrates speech-to-text processing, image-based medical analysis, and context-aware NLP using Groq Whisper and LLaMA-3.2 models. It achieves an average speech recognition accuracy of 91.3% and supports real-time response generation under 5 seconds. The proposed system enhances telemedicine accessibility, particularly for elderly and visually impaired users, while maintaining diagnostic relevance. Architecture, implementation, and preliminary evaluation are presented to demonstrate its impact and scalability.