arapgechina24-lgtm/nctirs-platform-v2

🛡️ National Cyber Threat Intelligence & Response System (NCTIRS) v2 — AI-Powered National Security & Smart Policing Intelligence Platform featuring Real-time Threat Analytics, SOAR Automation, Zero-Trust Architecture, and Google Gemini 2.0 Flash AI Engine. Built for Kenya 🇰🇪 # NCTIRS — National Cyber Threat Intelligence & Response System **Kenya's AI-Powered Cyber Fusion Center for Real-Time Threat Detection, Attribution, and Automated Response.** --- ## 🏛️ Executive Summary **NCTIRS** is a national-grade Cyber Threat Intelligence and Response System designed to protect Kenya's critical national infrastructure (CNI). Operating as a centralized Command and Control (C2) Fusion Center, it integrates multi-domain intelligence sources — including network telemetry, CCTV surveillance feeds, OSINT data, and historical threat geometry — leveraging Artificial Intelligence to predict, detect, attribute, and neutralize emerging cyber threats in real time. The platform is built on a three-layer architecture: - **Perception Layer** — IoT sensors, drone fleet, CCTV feeds, network sniffers - **Cognition Layer** — ML models, APT signature matching, threat classification (SENTINEL AI) - **Integrity Layer** — Blockchain ledger, DPA 2019 compliance, evidence chain-of-custody ## 🚀 Core Intelligence Capabilities ### 1. SENTINEL AI — Threat Analysis Engine (Google Gemini) AI-powered threat and incident analysis with MITRE ATT&CK mapping, Kenya-specific infrastructure context, and structured risk assessments. Falls back to rule-based analysis when offline. ### 2. AURA — Automated Universal Risk Attribution Python ML backend using Random Forest ensemble models to attribute threats to known APT groups (APT28, Lazarus, APT41, ZINC-24) based on behavioral telemetry patterns. ### 3. Sovereign AI — On-Premise LLM (Ollama) Local LLM capability via Ollama for threat analysis without external API dependency. Ensures digital sovereignty and data protection compliance under the Data Protection Act 2019. ### 4. Adversarial Defense & Federated Learning - Gradient masking, noise injection, adversarial training, ensemble voting - Federated learning across 24+ distributed nodes - Explainable AI (XAI) transparency panels ### 5. Real-Time NLP & OSINT Sentiment Ana …