Predict-Event: Real-Time AI-Powered Incident Intelligence Platform

NexEvent is an enterprise-grade, real-time event detection and threat intelligence platform designed to process, analyze, and visualize global incidents as they unfold. Leveraging advanced Natural Language Processing (NLP) and a high-performance modern web stack, it serves as a robust system for discovering critical events across multiple data streams with minimal latency.

Architecture Overview

Built for scale and real-time performance, NexEvent utilizes a decoupled client-server architecture:

1. Intelligence Engine (Backend)

• Framework: FastAPI (Python)
• Database: MongoDB for high-throughput, non-blocking asynchronous I/O
• Real-Time Communication: WebSockets for low-latency event broadcasting
• AI/ML Pipeline:
  • Named Entity Recognition (NER): Powered by spaCy to autonomously extract locations, organizations, and key figures from raw text streams.
  • Sentiment & Urgency Scoring: Utilizes VADER sentiment analysis combined with custom risk-scoring algorithms to assign actionable "Urgency Scores" (0-100) to incoming events.

2. Visualization & Command Center (Frontend)

• Framework: Next.js 14 (React) with App Router
• Language: TypeScript for strict type-safety and enhanced developer experience
• Styling & UI: Tailwind CSS for utility-first styling, integrated with a bespoke glassmorphic design system to ensure high-contrast readability under SOC (Security Operations Center) conditions.
• State Management: React Context API for centralized real-time stream state (EventContext) and Authentication (AuthContext).
• Data Visualization:
  • GPU-accelerated MapLibre GL JS integration for real-time geographic incident plotting.
  • recharts for dynamic, live-updating global sentiment and category velocity analytics.
  • framer-motion for complex staggered entry animations and seamless layout transitions.

Key Capabilities

• Asynchronous Streaming: Processes and broadcasts thousands of events per minute without blocking the main event loop.
• Smart Filtering & Search: O(1) complexity routing and highly-optimized client-side filtering by category, sentiment, and full-text keyword search.
• Automated Threat Detection: High-urgency events automatically trigger system-wide priority alerts and elevate to the "Critical Incidents" dashboard.
• Component Lazy-Loading: Next.js dynamic imports (next/dynamic) heavily utilized for heavy visualization libraries (Maps, Charts) to maintain optimal first contentful paint (FCP) and Time to Interactive (TTI).

Local Development Setup

Prerequisites

• Node.js (v18+)
• Python (3.12+)
• MongoDB instance (Local or Atlas)

Backend Deployment

cd backend
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python -m spacy download en_core_web_sm

# Configure environment variables (.env)
# Ensure MONGODB_URI is set

uvicorn app.main:app --host 0.0.0.0 --port 8001 --reload

Frontend Deployment

cd frontend
npm install

# Configure environment variables (.env.local)
# NEXT_PUBLIC_API_URL=http://localhost:8001
# NEXT_PUBLIC_WS_URL=ws://localhost:8001

npm run dev

Production Deployment

The application is container-ready and structurally designed to be deployed across modern cloud providers:

• Frontend: Recommend edge deployment via Vercel.
• Backend: Recommend Render, AWS ECS, or Google Cloud Run for containerized deployment.
• Database: MongoDB Atlas.

Ensure proper CORS configuration and secure connection protocols (WSS/HTTPS) for production environments.