A highly scalable, ultra-low-latency crisis management and dispatch platform designed for real-time emergency response. The system features a "God Mode" command center that provides live tracking of incidents, dynamic routing of first responders, instantaneous SOS broadcasting, and AI-driven tactical assessment using multimodal AI.
AI Integration: Google Gemini 2.5 Flash SDK (@google/genai)
Communication & Hosting: WhatsApp Meta API (Automated SOS Messaging), Oracle Cloud, Vercel
Routing: React Router v7
Key Features & What You Built
1. Live Map Integration & Real-Time Tracking (Command Center)
Engineered a highly interactive "God Mode" map interface capable of tracking moving entities with pinpoint accuracy.
Custom Mapbox Implementation: Integrated mapbox-gl to render a custom standard map style with hardware-accelerated rendering.
Live Entity Tracking: Mapped distinct visual markers for different entities (Police , Ambulance , Fire , Distressed Users ) that update in real-time as the database state changes.
Dynamic Geo-Routing: Leveraged the Mapbox Directions API to calculate and draw live, real-time routes on the map.
Spatial Calculations: Implemented Haversine formula logic to calculate exact distances and ETAs between emergencies and responders.
2. Insanely Fast Real-Time State with SpaceTimeDB
Integrated SpaceTimeDB to eliminate the need for traditional REST APIs and WebSockets, opting for a unified, ultra-fast data layer.
Live Tables: Connected React components directly to SpaceTimeDB tables using useTable.
Zero-Latency SOS Broadcast: Built the MinimalSOS component which instantly fires SpaceTimeDB reducers, registering a 5-Severity SOS and broadcasting the user as a live "Rescue Node" on the God Mode map.
3. Gemini Multimodal AI & Voice Integration
Built an "AI Dispatch Coordinator" by directly integrating Google's Gemini 2.5 Flash model to handle incoming crisis data intelligently.
Audio/Voice Processing: Programmed the system to accept audio/voice transcriptions of distress calls, feeding them directly into the Gemini model.
Tactical Data Extraction: Engineered precise prompts to force the AI to return structured JSON data, analyzing emergency severity and categorizing it.
Strategic Resource Allocation: Built a dispatch analyzer that evaluates incidents against a pool of available responders, prompting the AI to intelligently suggest resource deployment.
Crisis Image Verification: Added a visual layer where the AI verifies base64 image uploads to confirm if an image actually depicts a crisis with a confidence score.
4. Premium UI/UX & Tactical Aesthetics
Designed a premium, functional, and visually striking "analog-premium" dashboard.
Tactical Aesthetic: Utilized espresso and terracotta color palettes, tracking-widest uppercase typography, and high-contrast borders.
Glassmorphism & Fluidity: Used Framer Motion to create slick, slide-in side-panels (bg-white/85 backdrop-blur-xl) revealing deep data without page loads.
Micro-Animations: Implemented pulsing dots to visually indicate live traffic routing and active distress signals.
5. Automated WhatsApp Proximity Alerts
Integrated the Meta API to automatically reach out to users near an active emergency.
Proximity Messaging: When an incident is reported, the system automatically sends a WhatsApp message to users within the affected radius.
One-Tap Safety Verification: Users can simply tap 'Yes' or 'No' to indicate if they are safe.
SOS Location Sharing: If a user is not safe, they can instantly share their live location directly from WhatsApp to trigger a high-priority SOS alert on the God Mode Dispatch map.