AI-powered healthcare platform for intelligent medical consultations, symptom assessment, emergency triage, and patient health management.
Maritime Smart Care360 uses Large Language Models to analyze patient symptoms, estimate possible medical conditions, and recommend appropriate next steps — while maintaining secure healthcare data management. Built on a modern full-stack architecture, it combines conversational AI, electronic health record (EHR) management, and administrative tools into a single responsive web application.
⚠️ Medical Disclaimer This application is intended for educational and decision-support purposes only. It does not replace professional medical diagnosis or treatment. Always consult a qualified healthcare provider.
- Screenshots
- Features
- Technology Stack
- Architecture
- AI Workflow
- Database Modules
- Security
- Performance Optimizations
- Project Highlights
- Future Enhancements
- Disclaimer
The interface is shown in Japanese.
Quick access to AI consultation, doctor referral, patient management, and administrative tools.
Real-time AI doctor powered by Google Gemini, with conversational symptom analysis and suggested prompts.
Emergency severity level, possible diseases with confidence scores, a structured diagnosis summary, AI recommendations, and consultation metadata.
Automatic severity classification (Level 1–3) with emergency guidance and a condition-to-severity reference.
Real-time, streaming AI conversations powered by Google Gemini — natural-language symptom collection, context-aware follow-up questions, multi-turn dialogue, and built-in medical safety guidance.
Analyzes user-provided symptoms to generate possible conditions, confidence scores, symptom explanations, differential diagnosis suggestions, and personalized recommendations.
Evaluates symptom severity and assigns urgency levels (1–3) with emergency recommendations, hospital visit suggestions, hotline guidance, and clinical warning indicators.
Each consultation produces a structured report: primary symptoms, possible diseases, confidence percentages, severity assessment, recommended actions, and an AI medical summary.
Manage patient profiles, medical history, allergies, current medications, chronic conditions, and health records.
Appointment scheduling, upcoming appointments, history, visit records, and consultation notes.
Doctor directory, hospital search, specialty filtering, and healthcare provider information.
Every AI consultation is securely stored — chat history, diagnosis results, emergency level, recommendations, and timestamps.
User authentication, session management, protected routes, and secure API access.
User management, consultation monitoring, AI diagnosis history, patient records, healthcare analytics, and system management.
| Layer | Technologies |
|---|---|
| Frontend | React, TypeScript, Vite, Tailwind CSS, shadcn/ui, React Router, TanStack React Query |
| Backend | Supabase, PostgreSQL, Supabase Edge Functions |
| AI | Google Gemini, streaming AI responses, prompt engineering |
| Database | PostgreSQL, Row-Level Security (RLS), relational schema, patient data isolation |
| Auth | Supabase Auth, JWT authentication, protected APIs |
flowchart LR
Client["React Client"]
Backend["Supabase Backend<br/>Authentication · PostgreSQL · Edge Functions"]
Gemini["Google Gemini<br/>LLM API"]
Client -- "HTTPS / REST" --> Backend
Backend -- "Secure server-side API" --> Gemini
Gemini -- "AI response" --> Backend
Backend -- "Results" --> Client
flowchart TD
A["Patient describes symptoms"] --> B["Google Gemini analysis"]
B --> C["Symptom understanding"]
C --> D["Possible diseases"]
D --> E["Emergency severity classification"]
E --> F["Healthcare recommendations"]
F --> G["Consultation saved"]
Users · Patient Profiles · Medical History · Symptoms · Consultations ·
Diagnosis Results · Medications · Allergies · Appointments · Doctors · Hospitals ·
Chat History
Designed with healthcare data protection in mind:
- Row-Level Security (RLS) for per-user data isolation
- Secure authentication and protected API routes
- Server-side AI integration (API keys never exposed to the client)
- Encrypted communication
- TanStack React Query caching and background synchronization
- Optimistic UI updates
- Lazy loading and a component-based architecture
- Efficient API request management
- 🤖 AI-powered medical consultation with Google Gemini
- 🩺 Intelligent symptom analysis and disease prediction
- 🚨 Emergency severity classification and triage
- 📋 AI-generated diagnosis summaries
- 👤 Secure patient health record management
- 📅 Appointment scheduling and consultation history
- 🏥 Doctor and hospital discovery
- 👨💼 Administrative dashboard for monitoring consultations
- 🔒 Healthcare-grade data security using PostgreSQL Row-Level Security
- ⚡ Responsive React app with real-time streaming AI responses
- Voice-based AI consultation
- Medical image analysis
- Wearable device integration
- Prescription management
- Laboratory result tracking
- Multi-language support
- FHIR/HL7 interoperability
- Doctor-facing clinical dashboard
- Mobile application
- AI-assisted follow-up care
This application is intended for educational and decision-support purposes only and does not replace professional medical diagnosis or treatment. In an emergency, contact your local emergency services immediately.