MEDI-AI

MEDI is an AI-powered, voice-first healthcare communication assistant designed to help people organize and communicate their health concerns before speaking with a healthcare professional. Instead of requiring users to fill out complicated forms or type long descriptions, MEDI allows them to explain their concerns naturally through voice. Using real-time AssemblyAI voice technology, MEDI listens to the conversation, asks relevant follow-up questions, and organizes important information into structured patient context. MEDI can capture details such as the user's main concern, symptoms, duration, severity, location, associated symptoms, medications, allergies, medical history, and additional notes. Its structured patient-memory system allows information shared throughout the conversation to remain connected rather than being treated as isolated messages. MEDI is not designed to replace doctors, diagnose conditions, or prescribe medication. Its purpose is to improve communication and help users prepare clearer information for healthcare professionals. For potentially urgent situations, MEDI is designed to encourage appropriate professional or emergency care. Technically, MEDI combines a Python and FastAPI backend with WebSockets for real-time communication, AssemblyAI's voice AI, and an HTML/CSS/JavaScript frontend. The system demonstrates how conversational AI can go beyond simple speech-to-text by actively organizing information during a natural conversation. Our goal with MEDI is to make healthcare communication more accessible, natural, and organized. Instead of expecting people to know exactly what information to provide, MEDI lets them speak in their own words while AI helps structure the conversation. People should be able to explain what they're experiencing naturally, while technology helps organize the information—not get in the way.