Medi-K

Medi-K is a preventive health assistant that helps people register symptoms, track intensity, duration, sleep, mood, medication, and notes, then turns those records into an interactive health dashboard and a structured summary for medical consultation.

Instead of being only a text-based chatbot, Medi-K focuses on functional generative UI: the user can interact with forms, save symptom records, view calculated risk levels, monitor average intensity, delete records, and generate a consultation summary directly from the interface. The goal is not to diagnose or prescribe, but to help users organize health information clearly and avoid forgetting important details before speaking with a medical professional.

The project was built using the Generative UI Global Hackathon starter architecture, including a Next.js frontend, LangGraph agent environment, CopilotKit/AG-UI concepts, Gemini API integration, Notion integration, Docker infrastructure, PostgreSQL, and Redis. During the hackathon, we adapted the original lead-management concept into a health-prevention interface named Medi-K, redesigned the user experience with a medical blue visual style, and implemented a working symptom-recording dashboard.

Medi-K demonstrates originality by applying generative UI to preventive healthcare communication. It shows technical execution through working frontend code, local persistence, risk classification logic, and generated medical consultation summaries. The interface responds to user-provided data by updating the dashboard, classifying records, and creating structured output that can be copied and shared with a doctor.

This project was built during the hackathon using the provided Generative UI Global Hackathon starter kit as the base architecture. The starter kit originally included a lead-management workflow with Next.js, CopilotKit/AG-UI concepts, LangGraph, Notion, Docker, PostgreSQL, and Redis.

During the hackathon, our team configured the environment, fixed local setup issues, connected the infrastructure, and adapted the project into Medi-K, a preventive health assistant. The health-focused concept, visual identity, interface direction, symptom-recording workflow, risk classification logic, and consultation-summary generation were created during the hackathon.