Sentinel is an enterprise-grade AIOps platform designed to bridge the gap between development and operations. It implements a full DevSecOps Lifecycle—integrating automated testing, security vulnerability scanning, and AI-driven infrastructure self-healing.
- Automated Unit Testing: Integrated
Jestsuite for critical service verification (e.g., Authentication logic). - Security Scanning: Integrated Trivy vulnerability scanning for container images to detect CVEs before production.
- Deployment Governance: A "Locked-Gate" mechanism that prevents deployment if tests or security scans fail.
- Intelligent Observability: Real-time hardware telemetry via Prometheus and Node Exporter.
- Unsupervised ML: Uses Isolation Forest algorithms to detect system anomalies without manual threshold setting.
- AI Diagnosis: Leverages Llama-3 (Groq) to generate instant root-cause analysis and remediation bash commands.
- Automated Remediation: Direct integration with the Docker Engine API to restart failing workloads.
- Live Notifications: Real-time alerting via Telegram Bot for critical anomalies and "Green Ops" sustainability reports.
| Category | Tools |
|---|---|
| Frontend | React (Vite), Tailwind CSS, Chart.js |
| Backend | FastAPI (Python), Node.js (Express) |
| AI/ML | Scikit-learn (Isolation Forest), Llama-3 (Groq) |
| Infrastructure | Docker, Docker Compose, Prometheus, Node Exporter |
| Security/Test | Trivy, Jest (Auth Testing) |
| Communication | Telegram Bot API |
├── sentinel/ # Python AIOps Core (The "Brain")
│ ├── sentinel.py # FastAPI Backend & ML Logic
│ ├── sentinel_enterprise.html # DevSecOps Dashboard UI
│ └── .env # API Keys (Groq, Telegram)
├── rentacar-backend/ # Node.js API Service (The "Workload")
│ ├── server.js # Express Business Logic
│ └── tests/ # Jest Unit Tests (Auth_Test.js)
├── rentacar-frontend/ # Vite React Frontend (The "User Interface")
└── docker-compose.yml # Full Stack Orchestration
- Docker & Docker Compose installed.
- Groq API Key (for AI-powered Diagnosis).
- Telegram Bot Token & Chat ID (for Real-time Alerts).
-
Clone the repository:
git clone https://github.com/yourusername/sentinel-aiops.git cd sentinel-aiops -
Configure environment variables: Create a
.envfile inside thesentinel/directory:GROQ_API_KEY=your_key_here TELEGRAM_BOT_TOKEN=your_token_here TELEGRAM_CHAT_ID=your_id_here
-
Build and start the platform:
docker compose up -d --build
-
Access the dashboard: Open your browser and navigate to:
http://localhost:8000
Sentinel operates as a closed-loop autonomous system designed for high-availability environments:
-
Observability
- Prometheus scrapes real-time hardware and container metrics.
-
Detection
- An Isolation Forest machine learning model identifies statistical deviations from the learned baseline.
-
Diagnosis
- Groq AI (Llama-3) analyzes anomaly context and suggests remediation bash commands.
-
Action
- A human-in-the-loop triggers the "Execute AI Fix" button.
- The system communicates with the Docker Engine API to restart the affected service.
Sentinel includes a custom Green Ops Monitor.
If system load remains under 5% for more than 5 minutes, Sentinel will:
- Calculate wasted energy in kWh.
- Estimate the carbon footprint in gCO₂.
- Send a sustainability report to the administrator via Telegram.
This encourages:
- Resource decommissioning
- Infrastructure rightsizing
- Energy-efficient operations
Sentinel enforces a strict Shift-Left Security strategy.
Every code commit must pass through the DevSecOps Pipeline Guard:
-
Unit Testing
- Mandatory execution of the Jest test suite (e.g.,
Auth_Test.js).
- Mandatory execution of the Jest test suite (e.g.,
-
Vulnerability Scanning
- Automated Trivy scan of the container image.
-
Gatekeeping
- Deployment to production is programmatically locked unless:
- All tests pass.
- Security scans return a "Passed" status.
- Deployment to production is programmatically locked unless: