Architecture Overview

Aurora is a containerized application consisting of multiple services orchestrated via Docker Compose.

High-Level Architecture

┌─────────────────────────────────────────────────────────────────┐
│                         Frontend (Next.js)                       │
│                         http://localhost:3000                    │
└─────────────────────────────────────────────────────────────────┘
                                   │
                    ┌──────────────┴──────────────┐
                    ▼                              ▼
┌─────────────────────────────┐    ┌─────────────────────────────┐
│     REST API (Flask)        │    │   Chatbot (WebSocket)       │
│   http://localhost:5080     │    │   ws://localhost:5006       │
└─────────────────────────────┘    └─────────────────────────────┘
                    │                              │
                    └──────────────┬──────────────┘
                                   ▼
┌─────────────────────────────────────────────────────────────────┐
│                      Celery Workers                              │
│                   (Background Task Processing)                   │
└─────────────────────────────────────────────────────────────────┘
                                   │
        ┌─────────────┬────────────┼────────────┬─────────────┐
        ▼             ▼            ▼            ▼             ▼
┌─────────────┐ ┌─────────┐ ┌──────────┐ ┌─────────┐ ┌───────────┐
│  PostgreSQL │ │  Redis  │ │ Weaviate │ │  Vault  │ │ SeaweedFS │
│    :5432    │ │  :6379  │ │   :8080  │ │  :8200  │ │   :8333   │
└─────────────┘ └─────────┘ └──────────┘ └─────────┘ └───────────┘

Core Components

Frontend (Next.js)

• Technology: Next.js 15, TypeScript, Tailwind CSS, shadcn/ui
• Port: 3000
• Authentication: Auth.js
• Purpose: User interface for investigations, settings, and connectors

REST API (Flask)

• Technology: Flask, Python 3.11
• Port: 5080
• Entry point: server/main_compute.py
• Purpose: HTTP API for CRUD operations, authentication, connector management

Chatbot (WebSocket)

• Technology: Flask-SocketIO
• Port: 5006
• Entry point: server/main_chatbot.py
• Purpose: Real-time chat interface for natural language investigations

Celery Workers

• Technology: Celery with Redis broker
• Purpose: Background task processing for long-running operations
• Tasks: Investigation runs, connector sync, async operations

Data Stores

PostgreSQL

• Port: 5432
• Database: aurora_db
• Purpose: Primary data store for users, investigations, connector configs

Redis

• Port: 6379
• Purpose: Celery task queue, session cache, real-time subscriptions

Weaviate

• Port: 8080
• Purpose: Vector database for semantic search over logs and documents

HashiCorp Vault

• Port: 8200
• Purpose: Secrets management for user credentials and API keys

SeaweedFS

• Port: 8333 (S3 API), 8888 (File Browser)
• Purpose: S3-compatible object storage for file uploads

Agent Architecture

Aurora uses LangGraph for agent orchestration:

User Query
    │
    ▼
┌─────────────────────┐
│   Agent Supervisor  │
└─────────────────────┘
    │
    ├──► Cloud Connector Agents (GCP, AWS, Azure)
    ├──► Observability Agents (Datadog, Grafana)
    └──► Communication Agents (Slack, PagerDuty)

Each agent can:

• Query external systems using configured credentials
• Synthesize findings into natural language
• Chain together for complex investigations

Directory Structure

aurora/
├── client/              # Next.js frontend
│   └── src/
│       ├── app/         # App router pages
│       └── components/  # React components
├── server/              # Python backend
│   ├── routes/          # Flask blueprints
│   ├── connectors/      # Cloud/tool integrations
│   ├── agents/          # LangGraph agents
│   └── utils/           # Shared utilities
├── kubectl-agent/       # Kubernetes agent (optional)
└── website/             # Documentation (Docusaurus)

Docker Compose Files

File	Purpose
docker-compose.yaml	Development stack
docker-compose.prod-local.yml	Production-like testing and deployment