LagGuard

Real-time Network Prediction Overlay for Gamers & Streamers

A lightweight desktop overlay that predicts lag spikes before they happen using live network analysis and machine learning.

---

About

LagGuard is a predictive network monitoring overlay built for latency-sensitive applications like competitive gaming and live streaming.

Instead of showing lag after it happens, LagGuard continuously analyzes network behaviour in real time and warns users before instability becomes noticeable in-game.

The overlay is designed to stay lightweight, minimal, and distraction-free while remaining visible on top of fullscreen applications.

---

Preview

Warning Prediction

---

Stable Connection State

---

Compact Overlay Mode

---

Settings Panel

---

Theme Customization

---

Minimal Overlay UI

---

Features

• Real-time latency monitoring
• Jitter and packet loss analysis
• Predictive lag detection
• XGBoost-based ML predictions
• Frameless Electron overlay
• Click-through transparent UI
• WebSocket live communication
• Gaming optimized interface
• Lightweight background processing
• Configurable alerts and thresholds

---

Tech Stack

Layer	Technology
Frontend	Electron + HTML/CSS
Backend	Node.js
Machine Learning	XGBoost
Communication	WebSockets
Monitoring	ICMP / Ping Sampling

---

Architecture

 Network Sampling
        │
        ▼
 Feature Extraction
        │
        ▼
 Prediction Engine
 (Rules + XGBoost)
        │
        ▼
  WebSocket Stream
        │
        ▼
 Electron Overlay

---

Installation

Clone the repository

git clone https://github.com/yourusername/lagguard.git
cd lagguard

Install dependencies

npm install

Start development build

npm run dev

Build application

npm run build

---

Usage

Run the application:

npm start

Configure monitoring settings inside:

config.json

Example:

{
  "target": "8.8.8.8",
  "interval": 500,
  "predictionThreshold": 0.75
}

---

Prediction States

State	Meaning
Stable	Network healthy
Warning	Possible lag incoming
Danger	High probability lag spike

---

Roadmap

• Historical analytics dashboard
• Cloud synced profiles
• Game-specific presets
• Discord notifications
• Mobile companion app
• Multi-server monitoring
• Advanced ML training pipeline

---

Contributing

Pull requests are welcome.

If you'd like to contribute:

fork -> clone -> develop -> pull request

---

License

MIT License

---

If you like the project, consider starring the repository ⭐