🤖 Agentic AI for Autonomous Network Optimization

A multi-agent AI system for autonomous network monitoring and optimization using LLM-powered decision making.

🌐 Live Demo

🚀 Live Deployment: https://agentic-network-optimizeragentic-ai-for.onrender.com

Note: Free tier may take 30-50 seconds to wake up on first visit.

Run Locally:

git clone https://github.com/Ayan113/Agentic-AI-for-Autonomous-Network-Optimization.git
cd Agentic-AI-for-Autonomous-Network-Optimization
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
python run.py api
# Open http://localhost:8000 in your browser

---

🌟 Features

• Multi-Agent Architecture: Four specialized agents working in coordination

  • Monitor Agent: Collects network metrics, detects anomalies
  • Decision Agent: LLM-powered analysis and decision making
  • Action Agent: Executes corrective actions
  • Coordinator: Orchestrates the Monitor → Decide → Act loop

• Modern Web Dashboard: Real-time visualization with glassmorphism UI

• LLM-Based Decision Making: Uses GPT-4 or mock provider for intelligent analysis

• Network Simulation: Built-in simulator with configurable events and scenarios

• Feedback Loop: Learns from action outcomes to improve future decisions

• REST API: Full control via HTTP endpoints

🏗️ Architecture

┌─────────────────────────────────────────────────────────────┐
│                     COORDINATOR                              │
│   ┌──────────┐    ┌──────────┐    ┌──────────┐              │
│   │ MONITOR  │───▶│ DECISION │───▶│  ACTION  │              │
│   │  AGENT   │    │  AGENT   │    │  AGENT   │              │
│   └──────────┘    └──────────┘    └──────────┘              │
│        │              │    ▲           │                     │
│        ▼              ▼    │           ▼                     │
│   ┌──────────┐    ┌──────────┐    ┌──────────┐              │
│   │ NETWORK  │    │   LLM    │    │ FEEDBACK │              │
│   │SIMULATOR │    │ PROVIDER │    │   LOOP   │              │
│   └──────────┘    └──────────┘    └──────────┘              │
└─────────────────────────────────────────────────────────────┘

🚀 Quick Start

Installation

# Clone the repository
git clone https://github.com/Ayan113/Agentic-AI-for-Autonomous-Network-Optimization.git
cd Agentic-AI-for-Autonomous-Network-Optimization

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

Running the System

Web Dashboard (Recommended):

python run.py api --port 8000
# Open http://localhost:8000 in your browser

Demo Mode - Run a single optimization cycle:

python run.py demo

Continuous Monitoring:

python run.py monitor --interval 10 --cycles 5

🎮 Dashboard Features

Feature	Description
Health Ring	Real-time network health score (0-100)
Agent Status	Live status of all 4 agents
Network Metrics	Latency, bandwidth, packet loss, CPU, memory
Node Grid	Visual health status of all network nodes
Run Cycle	Trigger optimization on demand
Scenarios	Simulate network issues (outage, high traffic, etc.)

📊 API Endpoints

Endpoint	Method	Description
/	GET	Dashboard UI
/status	GET	System and agent status
/metrics	GET	Current network metrics
/decisions	GET	Decision history
/cycle	POST	Run optimization cycle
/simulate	POST	Trigger network scenario
/docs	GET	Interactive API documentation

⚙️ Configuration

Edit config.yaml to customize:

llm:
  provider: "mock"  # Options: "openai", "mock"
  model: "gpt-4o-mini"
  
network:
  simulation:
    enabled: true
    nodes: 10
    event_probability: 0.3

agents:
  monitor:
    polling_interval: 5
  decision:
    confidence_threshold: 0.7

Using OpenAI

export OPENAI_API_KEY="your-api-key"
# Then update config.yaml: provider: "openai"

📁 Project Structure

├── src/
│   ├── agents/           # AI Agents
│   │   ├── base_agent.py
│   │   ├── monitor_agent.py
│   │   ├── decision_agent.py
│   │   ├── action_agent.py
│   │   └── coordinator.py
│   ├── network/          # Network simulation
│   ├── llm/              # LLM integration
│   ├── feedback/         # Learning system
│   └── api/              # REST API
├── frontend/             # Web dashboard
│   ├── index.html
│   ├── styles.css
│   └── app.js
├── config.yaml           # Configuration
├── run.py               # Entry point
└── requirements.txt     # Dependencies

🔧 Available Actions

The system can execute these corrective actions:

Action	Description
optimize_routing	Optimize network paths
reduce_traffic	Throttle congested traffic
load_balance	Redistribute load
clear_cache	Free memory
restart_service	Restart problematic services
scale_up/down	Add/remove instances

📈 Performance

Tested results from demo:

• Health improvement: 93 → 97 (+4 points)
• Latency reduction: 29.7ms → 20.1ms (-32%)
• Packet loss reduction: 1.40% → 0.39% (-72%)

🤝 Contributing

Contributions welcome! Please feel free to submit a Pull Request.

📄 License

MIT License - feel free to use and modify for your projects.

👨‍💻 Author

Ayan Chatterjee

• GitHub: @Ayan113