Enhancing Rural Healthcare Accessibility Through AI-Driven Multilingual Symptom Triage On Low-End Smartphones

Abstract

Access to timely and accurate healthcare remains a persistent challenge in rural and semi-urban areas, predominantly where medical infrastructure is weak, professionals are scarce, and digital tools are inaccessible due to language or literacy barriers. This paper presents a lightweight, rule-based AI symptom triage system designed for deployment on low-end smartphones with support for regional languages such as Hindi and Kannada. The system allows users to report symptoms through a simple text-based interface and receive condition-based suggestions and guidance on whether home care or medical attention is needed. Developed using open-source platforms like Gradio, the tool focuses on usability and accessibility, with minimal computational requirements. A pilot deployment was conducted in a rural Indian setting, where the system received positive user feedback—especially regarding its language support, ease of use, and decision-making guidance. Preliminary results propose that such a system can help bridge the healthcare access gap by enabling informed self-triage. This study contributes to research on AI-assisted primary care systems in low-resource settings and lays the groundwork for future integration of voice support, machine learning-based diagnosis, and linkage with local healthcare providers.