🛡️ PulseSentry Health | Real-Time APM & Infrastructure Monitoring Platform

PulseSentry Health is a production-grade, real-time Application Performance Monitoring (APM) and server infrastructure surveillance platform built on the MERN stack. Designed with system engineering principles in mind, it provides low-latency telemetry streaming, threshold-based incident alerting, and Google Gemini AI-powered root cause analysis (RCA).

---

📸 UI Showcase

🌟 Landing Page (Light Mode)

🏥 Clinical Deck Dashboard (Light Mode)

📊 Hospital Management Operations (Dark Mode)

---

⚡ Key Architectural Capabilities

• 🚀 Low-Latency Real-Time Telemetry: Real-time bidirectional streaming of system metrics (CPU, RAM, Disk I/O, Network Throughput, Node.js event-loop lag) using a persistent WebSocket connection (socket.io).
• 🧠 AI-Powered Root Cause Analysis (RCA): Instantly diagnose system failures. PulseSentry integrates Google Gemini AI to analyze historical metric buffers, server event logs, and active process lists, outputting detailed diagnostic reports with remediation terminal commands.
• ⚙️ Dynamic Alerting & Rules Engine: A background rules evaluation loop that inspects incoming telemetry stream data against customizable threshold logic (e.g., CPU > 90% for 10s) and broadcasts push notifications instantly.
• 🖥️ Standalone Systems Monitoring Agent: A native agent script (agent/pulseAgent.js) using the systeminformation library. It can be run on any production machine to capture OS-level telemetry and stream it securely to your PulseSentry server.
• 🔄 Robust Zero-Config Fallbacks:
  • Database: Automatic fallback to a mock, in-memory store if MongoDB is unavailable, guaranteeing an instant out-of-the-box boot.
  • Gemini AI: Automatic fallback to simulated diagnostic templates if no Gemini API Key is configured.
• 🎭 Multi-Host Simulator: An in-app agent simulator control deck to trigger artificial memory leaks, CPU thrashing, or high network payloads across multiple virtual nodes to test alert engine compliance.

---

🏗️ System Architecture

graph TD
    subgraph Client Tier
        React[React Dashboard & Landing Page]
    end

    subgraph Backend API & Real-time Server
        Express[Express API Server]
        WSServer[Socket.io WebSocket Server]
        AlertEngine[Alerts Rules Evaluator]
    end

    subgraph Database Tier
        Mongo[(MongoDB / In-Memory Fallback)]
    end

    subgraph External Servers / Agents
        Agent1[PulseAgent - Production Server]
        Agent2[Simulated Agent - Web Server]
        Agent3[Simulated Agent - Database]
    end

    subgraph AI Tier
        Gemini[Google Gemini AI API]
    end

    Agent1 -- WebSocket Metrics --> WSServer
    Agent2 -- WebSocket Metrics --> WSServer
    Agent3 -- WebSocket Metrics --> WSServer

    WSServer -- Real-time Metrics Broadcast --> React
    Express -- REST APIs & Auth --> React
    React -- Request AI RCA --> Express
    Express -- Context Data --> Gemini
    Gemini -- Debugging Analysis --> Express
    Express -- Read/Write --> Mongo
    AlertEngine -- Background Rule Check --> Mongo
    AlertEngine -- Real-time Alert Broadcast --> WSServer

Loading

---

🛠️ Technology Stack

• Frontend: React 19, Vite, Tailwind CSS, Lucide Icons, Responsive Custom SVG Chart Engine.
• Backend: Node.js, Express, Socket.io (WebSocket), JWT Auth, bcryptjs.
• Database: MongoDB & Mongoose (with full in-memory fallback).
• AI Integration: @google/generative-ai (Gemini API).
• Agent: Node.js systems daemon using systeminformation & socket.io-client.

---

🚀 Getting Started

1. Prerequisites

• Node.js (v18+ recommended)
• MongoDB (Optional, the app runs on a mock database if MongoDB is not present)

2. Installation & Setup

Clone the repository and install all dependencies:

npm install

3. Environment Variables

Create a .env file in the root directory (the server will automatically load this):

PORT=5000
MONGODB_URI=mongodb://localhost:27017/pulsesentry
JWT_SECRET=your_jwt_secret_token_here
GEMINI_API_KEY=your_gemini_api_key_here

(Note: If GEMINI_API_KEY is left blank, the app will run with a simulated diagnostic fallback).

4. Running the Application

Start the frontend dev server and the backend API server concurrently with a single command:

npm run dev

Open http://localhost:5173 in your browser.

5. Running the Telemetry Agent

To monitor your actual local computer or a remote server:

1. Open a new terminal window.
2. Navigate to the project folder.
3. Run the standalone agent script:

node agent/pulseAgent.js

The agent will establish a WebSocket handshake with the server, auto-register your hostname, and start streaming live system telemetry directly to the dashboard!

---

📈 Portfolio Highlights

This project was built to demonstrate:

1. Concurrency: Seamless management of concurrent real-time client UI connections and background telemetry agent streams on a single Node server.
2. Resilience: Designing applications that fail gracefully (dynamic database fallback & mock-AI integration), making them highly reliable for recruiters and evaluators.
3. Advanced Networking: Direct manipulation of WebSockets using socket.io for high-frequency time-series datasets.
4. Clean Code: High-fidelity React hooks, customized modular page layouts, and isolated REST routers.