🚨 ResQAI

AI-Powered Disaster Intelligence & Emergency Response Platform

Predict. Coordinate. Respond. Save Lives.

ResQAI is a production-grade, full-stack disaster response and crisis intelligence platform. It unifies emergency distress reporting, real-time spatial GIS mapping, tactical rescue squad dispatching, relief shelter inventory management, volunteer mission tasking, executive analytics, and grounded AI decision support into a single high-reliability command center.

🌐 Try Live Application · ⚙️ Backend API · 📚 API Documentation · 💻 GitHub Repo

📌 Project Snapshot

Attribute	Specification
Core Concept	Integrated AI-assisted disaster management command center for India & high-risk regions
Primary Audience	Incident Commanders, NDRF/SDRF Rescue Squads, District Collectors, Relief NGOs, Affected Citizens
Core Value	Replaces fragmented phone calls and paper logs with real-time GIS telemetry and grounded AI triage
Key Differentiator	Full-stack production readiness, 6-role RBAC, real PostgreSQL database, and TEE-ready confidential computing architecture

🌐 Live Deployments

Resource	Link	Description
🌐 Live Web Application	https://resqai-kappa.vercel.app	Production Single-Page React SPA Client
⚙️ Backend API Server	https://resqai-backend.vercel.app	Standalone FastAPI Serverless API on Vercel
📚 Swagger API Documentation	https://resqai-backend.vercel.app/docs	Interactive OpenAPI 3.1.0 Endpoint Explorer
💻 Source Repository	https://github.com/mrashish18/ResQAI	GitHub Source Code & Issue Tracker

⚠️ The Problem vs 💡 The ResQAI Solution

Structural Crisis Bottleneck	ResQAI Solution Engine
1. Fragmented Communication & Silos Affected citizens struggle to locate open shelters, while first responders and NGOs lack visibility into local crisis severity.	Unified Operations Platform Single dashboard aggregating citizens, rescue teams, NGOs, and government command into one shared operational view.
2. Delayed Emergency Extraction Rescue squads operate with incomplete or outdated coordinates, prolonging response times for critical SOS distress signals.	Real-Time GIS Operations Map Interactive Leaflet dark basemap displaying live GPS distress coordinates, squad positions, and priority markers.
3. Imbalanced Relief Distribution Food, medical supplies, and clean water are allocated using guesswork, leading to inventory hoarding or shortages.	Shelter & Resource Inventory Manager Real-time occupancy tracking, itemized warehouse inventory, and automated distribution logging against specific incidents.
4. Uncoordinated Volunteer Mobilization Willing community volunteers lack a transparent platform to view, accept, and update critical field tasks.	Volunteer Mission Task Board Public task board allowing volunteers to browse priority assignments, accept tasks, and update completion status.
5. Cognitive Overload for Officials Command officials face massive streams of unorganized field telemetry, making it difficult to identify priorities.	AI Command Assistant & Analytics Telemetry-grounded AI assistant (OpenRouter) providing instant database queries and Chart.js crisis analytics.

🏗️ Project Architecture

🌟 Key Capabilities

Domain	Core Capability	Implementation Details
🧠 AI Command Center	Telemetry-Grounded AI Assistant	Answers complex crisis queries using live database context (incidents, shelters, squads) with verifiable source citations via OpenRouter API.
🗺️ Live Emergency Map	Spatial Operations GIS Map	Custom Leaflet dark basemap featuring real-time incident markers, shelter capacity badges, rescue squad tracking, and risk heatmaps.
🚨 SOS Broadcasts	Priority Emergency Distress	Enables citizens to broadcast emergency distress signals with exact GPS coordinates, medical flags, victim counts, and image uploads.
🚑 Rescue Squad Dispatch	Tactical Unit Coordination	Coordinates specialized units (NDRF, SDRF, Coast Guard, Army Engineers), assignment tracking, and status lifecycle transitions.
🏠 Shelters & Relief	Inventory & Capacity Tracking	Tracks real-time shelter occupancy, warehouse stock levels (food, water, medicine), and logs itemized supply distributions.
👥 Volunteer Task Board	Field Mission Coordination	Transparent task board for community volunteers to browse priority assignments, accept tasks, and submit completion status.
📊 Crisis Analytics	Executive Decision Dashboard	Visualizes incident severity distributions, resource allocation ratios, and temporal crisis trends using Chart.js.
⚙️ System Settings	Operations Control Center	Multi-domain settings for theme density, 8 dispatch alert channels, quiet hours, GIS basemaps, offline storage sync, and health checks.
🔐 Security & RBAC	Enterprise Security & TEE-Ready	JWT authentication, 6 role-based access rules (Citizen, Volunteer, Rescue, NGO, Government, Admin), and TEE-ready confidential compute pipeline.

📸 Product Walkthrough (30 Screenshots)

🌐 1. Platform & Authentication (Screenshots 01–09)

Interface	Preview
Landing Hero Page
Platform Feature Overview
AI Hazard Prediction Engine
Complete Disaster Operations
First Responder Platform
Active Disaster Response Telemetry
Community Protection Architecture
User Account Registration Console
Secure Authentication Portal

🗺️ 2. Emergency Operations & GIS (Screenshots 10–14)

Interface	Preview
Operational Command Dashboard
Live Emergency Operations Map
Live Map Spatial Telemetry
GIS Satellite & Basemap Layers
Citizen SOS Emergency Broadcast

🚁 3. Rescue, Shelters & Resources (Screenshots 15–21)

Interface	Preview
Rescue Team Squad Operations
Register Rescue Squad Modal
Shelters & Resource Inventory
Relief Shelter Management
Register Relief Shelter Modal
Register Inventory Item Modal
Volunteer Mission Task Board

🤖 4. AI & Analytics (Screenshots 22–25)

Interface	Preview
ResQAI Intelligence Assistant
AI Command Center Chat Console
Automated AI Risk Assessment
Crisis Analytics & Telemetry Insights

⚙️ 5. System, Security & Administration (Screenshots 26–30)

Interface	Preview
Emergency Broadcasts & Alerts Center
System Settings Control Center
Dispatch Alert Channels & Settings
Admin Security Console
Provision User Account Modal

🛠️ Technology Stack

Layer	Component Technologies
Frontend SPA	React 19.2, TypeScript 5.0, Vite 8.1, Tailwind CSS v4.0, Framer Motion
GIS & Spatial Mapping	Leaflet 1.9, React Leaflet 5.0, CartoDB Voyager / Dark Basemaps, Esri Topo / Satellite
Data Visualization	Chart.js 4.4, react-chartjs-2, Lucide React Icons
Backend Framework	FastAPI 0.111, Uvicorn 0.30 (ASGI), Pydantic v2, Python 3.11 / 3.12
Database & ORM	PostgreSQL 15 (Production / Neon), SQLite (Local Dev), SQLAlchemy 2.0, Alembic 1.13
Security & Auth	Python-Jose (JWT), Passlib (Bcrypt), OAuth2 Password Bearer, 6-Tier RBAC
AI Intelligence	OpenRouter API (`nvidia/llama-nemotron-rerank-vl-1b-v2:free`), HTTPX Async Client
Deployment	Vercel Serverless Functions (`@vercel/python`), Vercel Static Build (Frontend SPA), Docker Compose

🔐 Confidential Computing & TEE Architecture (TEE-Ready)

ResQAI features a TEE-ready Confidential Computing architecture designed to protect sensitive emergency workloads (such as citizen SOS distress calls, victim locations, and tactical triage recommendations) when deployed on supported hardware enclaves.

User Payload (SOS / Victim Data)
  │
  ▼
FastAPI Security Gateway
  │
  ▼
sanitize_and_minimize_input()  <-- Scrubs PII, emails, phone numbers & credentials
  │
  ▼
ConfidentialComputeService (Provider Abstraction)
  │
  ├── [CONFIDENTIAL_COMPUTING_ENABLED=true]  ──> AWS Nitro / GCP Confidential / Azure TEE Enclave
  │
  └── [CONFIDENTIAL_COMPUTING_ENABLED=false] ──> Development Fallback Provider
                                                  (Truthfully reports: "Development Mode — TEE Off")

Security Architecture Rules & Status

Data Minimization Pipeline: sanitize_and_minimize_input() automatically scrubs PII, credentials, and non-essential fields prior to enclave processing.
Provider-Agnostic Abstraction: Designed for deployment pathways on AWS Nitro Enclaves, GCP Confidential Space, and Azure Confidential VMs.
Truthful Status Reporting: In development / serverless mode (CONFIDENTIAL_COMPUTING_ENABLED=false), the system explicitly labels processing as ⚠️ Development Mode — Hardware TEE Unavailable without making false security claims.
Status Verification Endpoint: GET /api/security/confidential-computing.

For complete technical specifications, see docs/confidential-computing.md.

🔑 Operational Access Roles

Role	Email Identifier	Demo Password	Access Level & Scope
Admin	`admin@resqai.in`	Not published	Full system control, user provisioning, task assignment & resource allocation
Government Official	`govt@resqai.in`	Not published	High-level crisis analytics, incident oversight & resource distribution approval
NGO Coordinator	`ngo@resqai.in`	Not published	Shelter management, warehouse inventory tracking & relief distribution
Rescue Squad Lead	`rescue@resqai.in`	Not published	Rescue squad status updates, incident response & victim extraction
Volunteer	`volunteer@resqai.in`	Not published	Volunteer task board, mission acceptance & field completion reporting
Citizen	`citizen@resqai.in`	Not published	Emergency SOS broadcast, shelter lookup & incident distress reporting

Note: Demo account credentials for hackathon evaluation are configured in the evaluation environment.

⚡ Quick Start & Installation

Prerequisites

Node.js v20.0+
Python 3.11+
Git

1. Clone Repository

git clone https://github.com/mrashish18/ResQAI.git
cd resqai

2. Backend Setup

cd backend

# Create & activate virtual environment
python -m venv .venv

# Windows:
.venv\Scripts\activate
# macOS / Linux:
source .venv/bin/activate

# Install backend dependencies
pip install -r requirements.txt

3. Frontend Setup

cd ../frontend
npm install

4. Run Development Servers

Start FastAPI Backend:

# In backend directory (venv active):
python -m uvicorn main:app --port 8000 --reload

API Base: http://localhost:8000
Interactive OpenAPI Docs: http://localhost:8000/docs
Health Check: http://localhost:8000/health

Start React Frontend:

# In frontend directory:
npm run dev

Local Client App: http://localhost:5173

📡 API Overview

The ResQAI backend exposes modular REST API controllers. Interactive documentation is available live at https://resqai-backend.vercel.app/docs.

Controller	Base Path	Description
Auth & Users	`/api/auth`, `/api/users`	JWT authentication, registration, profile updates, and RBAC user provisioning
Incidents	`/api/incidents`	Incident lifecycle, geographic coordinates, severity, and status transitions
Emergency SOS	`/api/sos`	Citizen distress broadcasts, medical emergency flags, victim counts, and image uploads
Shelters	`/api/shelters`	Shelter registration, capacity tracking, current occupancy, and status management
Rescue Teams	`/api/rescue`	Tactical rescue squad roster, dispatch tracking, and incident assignments
Resources & Relief	`/api/resources`	Supply inventory tracking and itemized disaster distribution logging
Volunteers	`/api/volunteers`	Field mission task board, task creation, volunteer acceptance, and completion
AI Command Center	`/api/chatbot`	Grounded AI decision support querying active database telemetry
Hazard Predictions	`/api/predictions`	Automated disaster risk prediction models and regional risk assessments
Crisis Analytics	`/api/analytics`	Executive summary stats, incident trends, resource ratios, and telemetry insights
Security & TEE	`/api/security`	Confidential computing status verification and sanitized enclave processing

🧪 Testing & Verification

Backend Integration Tests (38/38 Passing)

Run the automated end-to-end integration test suite:

python backend/test_api.py

RESQAI BACKEND — API TEST RESULTS
============================================================
[PASS] GET / -> 200 (expected 200)
[PASS] GET /health -> 200 (expected 200)
...
[PASS] GET /api/security/confidential-computing -> 200 (expected 200)
[PASS] POST /api/security/process-confidential-ai -> 200 (expected 200)
============================================================
38/38 tests passed — All tests passed!

Frontend Production Build

Validate TypeScript compilation and Vite bundling:

cd frontend
npm run build

📁 Repository Structure

resqai/
├── .env.example
├── .gitignore
├── docker-compose.yml
├── package.json
├── package-lock.json
├── README.md
│
├── backend/
│   ├── alembic/
│   ├── routes/
│   │   ├── analytics.py
│   │   ├── auth.py
│   │   ├── chatbot.py
│   │   ├── incidents.py
│   │   ├── notifications.py
│   │   ├── predictions.py
│   │   ├── rescue.py
│   │   ├── resources.py
│   │   ├── security.py
│   │   ├── shelters.py
│   │   ├── sos.py
│   │   ├── uploads.py
│   │   ├── users.py
│   │   ├── volunteers.py
│   │   └── weather.py
│   │
│   ├── services/
│   │   ├── ai.py
│   │   └── confidential_compute.py
│   │
│   ├── alembic.ini
│   ├── auth.py
│   ├── config.py
│   ├── database.py
│   ├── Dockerfile
│   ├── main.py
│   ├── models.py
│   ├── requirements.txt
│   ├── schemas.py
│   ├── seed.py
│   ├── test_api.py
│   └── utils.py
│
├── frontend/
│   ├── src/
│   │   ├── components/
│   │   ├── context/
│   │   ├── lib/
│   │   ├── pages/
│   │   ├── types/
│   │   ├── App.tsx
│   │   └── index.css
│   │
│   ├── Dockerfile
│   ├── index.html
│   ├── nginx.conf
│   ├── package.json
│   ├── package-lock.json
│   └── vite.config.ts
│
├── docs/
│   └── confidential-computing.md
│
└── screenshots/                    # 30 project screenshots

🏆 Impact & Project Differentiators

High-Impact Outcomes

Accelerated Emergency Response: Eliminates communication bottlenecks by directly connecting citizen SOS broadcasts to nearest rescue teams and GIS maps.
Resource Equity: Prevents relief inventory hoarding by providing real-time stock and occupancy visibility to relief coordinators.
Enhanced Command Intelligence: Reduces cognitive overload for incident commanders via grounded AI decision support and real-time Chart.js analytics.

Why ResQAI?

Production-Grade Full-Stack Architecture: Clean separation between React 19 SPA frontend and FastAPI ASGI backend.
Real Database Integration: Built on SQLAlchemy 2.0 with PostgreSQL 15 / Neon compatibility and zero hardcoded mock data.
Interactive Spatial GIS: Multi-basemap Leaflet engine with custom cluster markers, risk zones, and live telemetry fly-to controls.
Grounded AI Intelligence: LLM integration grounded strictly against live database tables with source citations.
Confidential Computing Readiness: Architected with a PII minimization pipeline and TEE abstraction layer for hardware enclave execution.

🚀 Phased Roadmap & Future Improvements

🟢 Phase 1 — Current Production Release (Implemented)

Full-Stack Crisis Operations: Incident lifecycle management, priority SOS distress broadcasting, and volunteer mission task board.
Interactive GIS Mapping: Multi-basemap Leaflet engine with custom cluster markers, risk zone overlays, and location fly-to controls.
Shelter & Resource Inventory: Real-time shelter capacity tracking, warehouse stock management, and itemized relief distribution logging.
Grounded AI Command Center: AI assistant (OpenRouter) grounded directly against live database telemetry with source citations.
Enterprise Security & TEE-Ready: 6-role RBAC, JWT authentication, PII minimization pipeline, and TEE confidential compute abstraction.

🟡 Phase 2 — Near-Term Platform Enhancements (Planned)

📡 WebSockets Real-Time Telemetry: Upgrade live map telemetry from HTTP polling to WebSocket streams for sub-second rescue squad GPS updates.
📟 IoT Sensor & Telemetry Integration: Ingest real-time field telemetry from water level sensors, weather stations, and seismic monitors.
🌦️ Official Weather & Hazard Feeds: Direct API integrations with official meteorological and disaster alert feeds for automated hazard triggers.
📡 Offline Mesh Networking: WebRTC peer-to-peer mesh fallback for emergency distress reporting in low-connectivity disaster zones.

🔵 Phase 3 — Long-Term Vision (Planned)

🛰️ Satellite & Drone Imagery Analysis: Aerial drone reconnaissance and satellite computer-vision pipelines for automated flood zone detection and structural damage assessment.
🚑 AI Rescue-Route & Evacuation Optimization: Dynamic predictive evacuation planning and real-time optimal routing for emergency extraction squads.
🗣️ Multilingual Voice Emergency Assistant: Multi-dialect voice assistant for hands-free SOS reporting in regional languages.
📱 Dedicated Native Mobile Apps: Cross-platform iOS/Android applications for citizens and first responders with offline-first sync.

📜 License & Intellectual Property

ResQAI is developed and published for portfolio, educational, demonstration, and hackathon evaluation purposes.

Unless otherwise stated, the project's original source code, design, documentation, visual assets, branding, and other original project materials are the property of the project authors.

Third-party libraries, frameworks, APIs, map data, icons, fonts, and other external resources remain subject to their respective licenses and terms of use.

Please do not reproduce, redistribute, or commercially reuse the project's original materials without permission from the project authors.

ResQAI — AI-Powered Disaster Intelligence & Emergency Response

🌐 Live Application · ⚙️ Backend API · 📚 API Documentation

Built with ❤️ for saving lives · ResQAI 2026

Navigation Menu

Repository files navigation

🚨 ResQAI

AI-Powered Disaster Intelligence & Emergency Response Platform

📌 Project Snapshot

🌐 Live Deployments

⚠️ The Problem vs 💡 The ResQAI Solution

🏗️ Project Architecture

🌟 Key Capabilities

📸 Product Walkthrough (30 Screenshots)

🛠️ Technology Stack

🔐 Confidential Computing & TEE Architecture (TEE-Ready)

Security Architecture Rules & Status

🔑 Operational Access Roles

⚡ Quick Start & Installation

Prerequisites

1. Clone Repository

2. Backend Setup

3. Frontend Setup

4. Run Development Servers

📡 API Overview

🧪 Testing & Verification

Backend Integration Tests (38/38 Passing)

Frontend Production Build

📁 Repository Structure

🏆 Impact & Project Differentiators

High-Impact Outcomes

Why ResQAI?

🚀 Phased Roadmap & Future Improvements

🟢 Phase 1 — Current Production Release (Implemented)

🟡 Phase 2 — Near-Term Platform Enhancements (Planned)

🔵 Phase 3 — Long-Term Vision (Planned)

📜 License & Intellectual Property

About

Releases

Contributors

Languages

Name		Name	Last commit message	Last commit date
api		api
backend		backend
docs		docs
frontend		frontend
screenshots		screenshots
.env.example		.env.example
.gitignore		.gitignore
README.md		README.md
docker-compose.yml		docker-compose.yml
package-lock.json		package-lock.json
package.json		package.json
vercel.json		vercel.json

Latest commit

History

Folders and files

Repository files navigation

🚨 ResQAI

AI-Powered Disaster Intelligence & Emergency Response Platform

📌 Project Snapshot

🌐 Live Deployments

⚠️ The Problem vs 💡 The ResQAI Solution

🏗️ Project Architecture

🌟 Key Capabilities

📸 Product Walkthrough (30 Screenshots)

🛠️ Technology Stack

🔐 Confidential Computing & TEE Architecture (TEE-Ready)

Security Architecture Rules & Status

🔑 Operational Access Roles

⚡ Quick Start & Installation

Prerequisites

1. Clone Repository

2. Backend Setup

3. Frontend Setup

4. Run Development Servers

📡 API Overview

🧪 Testing & Verification

Backend Integration Tests (38/38 Passing)

Frontend Production Build

📁 Repository Structure

🏆 Impact & Project Differentiators

High-Impact Outcomes

Why ResQAI?

🚀 Phased Roadmap & Future Improvements

🟢 Phase 1 — Current Production Release (Implemented)

🟡 Phase 2 — Near-Term Platform Enhancements (Planned)

🔵 Phase 3 — Long-Term Vision (Planned)

📜 License & Intellectual Property

About

Topics

Resources

Stars

Watchers

Forks

Releases

Contributors

Languages