🚨 ResQMe AI Responder Dashboard & SOS System

An emergency response system built for RiverHacks 2025, combining AI triage, real-time location context, and message-based SOS triggering.
Users can create their profiles via web interface, trigger SOS through iMessage (LoopMessage), and responders can monitor incidents via an intuitive dashboard.

---

🛠 Project Overview

This project enables users to pre-create emergency profiles and allows SOS alerts to be triggered quickly using SMS/iMessage. The system:

• Collects user health and emergency details
• Integrates with a backend server to handle incoming SOS triggers
• Fetches nearby hospitals, police stations, fire stations, and weather conditions based on user's live location
• Displays active SOS alerts on a Responder Dashboard
• Uses an LLM-based system for triage suggestions and severity prediction

The goal: rapid, context-aware emergency assistance with minimal user effort.

---

🧩 Tech Stack

Layer	Technology
Frontend	Next.js, TailwindCSS, TypeScript
Backend	Node.js (Express.js), Prisma ORM, TypeScript
Database	PostgreSQL
AI/Triage	Local LLM API (Navigator by WebAI)
Messaging	LoopMessage (for iMessage/SMS SOS triggers)
Deployment	Render / Vercel (can be self-hosted too)
External APIs	SERP API (for Nearby Context)

---

🚀 Features

1. Profile Management

• Create and update user profiles with:
  • Medical conditions
  • Emergency contacts
  • Communication preferences
  • Device info
• Consent for live location tracking.

2. SOS System

• Trigger SOS via iMessage (LoopMessage).
• Server listens for incoming messages and matches registered users.
• If no location is found, prompts the user to share.

3. Responder Dashboard

• Web dashboard listing all active SOS alerts.
• Displays:
  • Severity (Critical, Moderate, Minor)
  • Last known location
  • Profile details
  • Responder status (Accepted / On the way / Resolved)

4. Nearby Context Fetching

• Auto-fetch nearby hospitals, police stations, fire stations, and current weather based on user's location.
• Enriches responder decision-making.

5. LLM-Based Triage

• Health of LLM API is monitored.
• LLM processes SOS message content to:
  • Predict severity
  • Suggest next best action
  • Provide summary reasoning.

---

🧩 Mini Apps We Built

App Name	Description	Link
NearByMe	A Node.js app that fetches nearby hospitals, police stations, fire stations, and weather using geolocation and the SERP API.	http://localhost:3232
Responder Dashboard	A web-based dashboard for responders to monitor live SOS alerts, view user profiles, and manage response statuses intuitively.	http://localhost:3002/
SOSsystem	The core backend system that processes incoming SOS triggers, handles location requests, triages alerts with LLM support, and manages alert routing.	Runs internally on server
User Profile	A web app for users to create and update their emergency profiles including medical history, emergency contacts, and device information.	http://localhost:3000/profile
WebAI Navigator API	A backend system hosting the local LLM model (Navigator/Ollama), providing APIs for AI-powered triage, severity prediction, and emergency reasoning.	http://localhost:10501
LoopMessage SOS Trigger	External system to trigger SOS by sending a message.	Send iMessage to: sandbox.loopmessage.com@imsg.im

---

🗺️ Mini Apps Connection Diagram

┌─────────────────────┐
│     User Profile     │
│ (Next.js Web App)    │
└─────────┬───────────┘
          │ Profile Created / Updated
          ▼
┌─────────────────────┐        ┌─────────────────────┐
│      SOSsystem       │◀──────│   SOS Trigger via    │
│ (Node.js Backend)    │        │   LoopMessage (iMsg) │
└─────────┬───────────┘        └─────────────────────┘
          │
          ▼
┌─────────────────────┐
│      NearByMe        │
│ (Node.js Mini-App)   │
│ Fetch Nearby Context │
└─────────┬───────────┘
          │
          ▼
┌─────────────────────┐
│ WebAI Navigator API  │
│ (LLM Triage Engine)  │
│ Severity Prediction  │
└─────────┬───────────┘
          │
          ▼
┌─────────────────────┐
│  Responder Dashboard │
│ (Next.js Web Frontend)│
│ View & Manage Alerts │
└─────────────────────┘

---

🌟 Future Work / Extensions

• End-to-End Encryption: Secure SOS messages and user profiles.
• Responder Mobile App: Lightweight mobile app for responders to get push notifications.
• Multi-step Triage: Collect ongoing status updates from the user after SOS.
• Multi-language Support: Support for Spanish, Nepali, etc., for broader accessibility.
• Automated Location Tracking: Use device sensors/API integration for real-time movement tracking post-SOS.
• Incident Reporting Analytics: Post-incident analytics for responders and city authorities.
• Decentralized Hosting: Explore using P2P decentralized hosting for critical infrastructure resilience.

👥 Team Members

• Amul Poudel
  GitHub Profile

• Samar Ranjit
  GitHub Profile

• Gaurab Chhetri
  GitHub Profile

• Dipesh Pandit Github Profile

• Diwas Pandit Github Profile