انتقل إلى المحتوى الرئيسي

Welcome to AuroraSOC

AuroraSOC is an enterprise-grade, decentralized Agentic AI Security Operations Center (SOC) that combines cutting-edge AI agent technology with hardware-rooted IoT/CPS security. It is built on the IBM BeeAI Agent Framework and designed to automate the full lifecycle of security operations — from alert triage to incident response.

What is AuroraSOC?

Think of AuroraSOC as a team of 16 AI security specialists working together to protect your organization. Each agent is an expert in a specific domain — just like a real SOC team — but powered by Large Language Models (LLMs) and connected to real security tools.

Why AuroraSOC?

Traditional SOCs face several challenges:

ChallengeTraditional SOCAuroraSOC
Alert FatigueAnalysts manually triage thousands of alertsAI agents auto-triage, deduplicate, and correlate alerts
Slow ResponseHours to investigate and respondMinutes through automated playbook execution
IoT/CPS Blind SpotLimited visibility into OT/IoT devicesHardware-rooted attestation with physical-cyber correlation
Knowledge SilosExpertise trapped in individual analystsShared three-tier memory system across all agents
ScalabilityHiring more analysts is expensiveDeploy more agent replicas on demand
24/7 CoverageShift rotations with human fatigueAI agents operate continuously without fatigue

Key Design Principles

AuroraSOC is built on these foundational principles:

  1. No Kafka, No LangChain, No LangGraph — Uses Redis Streams + NATS JetStream for event bus and IBM BeeAI for agents. This avoids the complexity and overhead of Kafka while providing all necessary event streaming capabilities, and uses a purpose-built agent framework instead of generic LLM chains.

  2. Hardware-Rooted Trust — Every IoT/CPS device performs cryptographic attestation using hardware security modules, ensuring firmware integrity from the physical layer up.

  3. Human-in-the-Loop — High-risk actions always require human approval. The AI assists, but humans remain in control of critical decisions.

  4. Graceful Degradation — If the database is down, the API serves demo data. If an agent fails, the circuit breaker pattern prevents cascading failures.

  5. Full Observability — Every agent action is traced with OpenTelemetry, logged with structured JSON, and metriced with Prometheus.

Technology Stack at a Glance

Who is This Documentation For?

This documentation is organized into two main sections:

📖 User Guide (You Are Here)

For SOC analysts, operators, and administrators who use AuroraSOC daily:

  • How to navigate the dashboard
  • How to manage alerts and cases
  • How to configure and monitor AI agents
  • How to manage CPS/IoT devices
  • Understanding the security model

🛠️ Developer Guide

For developers and engineers who build, extend, or maintain AuroraSOC:

  • System architecture deep-dives
  • How each component is implemented and why
  • How to add new agents, tools, and integrations
  • API reference and testing guide

Quick Navigation

I want to...Go to...
Get AuroraSOC running quicklyQuick Start
Understand the dashboardDashboard Overview
Learn about the AI agentsAgentic AI SOC Concepts
Manage alerts and casesAlert Management
Understand authenticationAuthentication
Monitor CPS/IoT devicesCPS/IoT Devices