SentinelCore

SentinelCore is an enterprise-grade telemetry ingestion, processing, and predictive analytics platform. It collects high-throughput system event logs from Windows nodes, buffers them through Kafka, routes them into an isolated PostgreSQL warehouse, and visualizes system health via a sleek React dashboard backed by real-time FastAPI endpoints and predictive ML forecasting.

Architecture Highlights

• Distributed Ingestion: Windows collectors securely publish event telemetry to a localized Kafka cluster using an injectible pre-shared key (PSK).
• Asynchronous Persistence: kafka_to_postgres.py offloads high-throughput streams into a PostgreSQL database with dynamic caching, fallback circuit-breaking, and partitioned schema scaling.
• Predictive ML Analytics: ml_engine.py processes live feature_snapshots to passively compute system anomaly scores and hardware failure probabilities.
• Zero-Trust Dashboard: Real-time metrics are accessible only through a Google Firebase authenticated React frontend, which pulls normalized payloads from a token-restricted api_server.py.

Documentation Hub

For an exhaustive breakdown of SentinelCore's architecture and capabilities, explore our detailed documentation:

1. Architectural Overview - System design and data flow layout.
2. Backend Services Overview - Detailed behavior of Python microservices.
3. Frontend Dashboard - React UI configurations and context state mapping.
4. API Reference - Payload strictures for the FastAPI endpoints.

Infrastructure & Deployment Guides

• WSL / Local Deployment
• Firebase Authorization Configuration
• Server Network Security (UFW)

Quick Start (Backend)

Requirements: Python 3.9+, Kafka, PostgreSQL

# Terminal 1: Telemetry Consumer
python src/kafka_to_postgres.py

# Terminal 2: ML Snapshot Engines
python src/feature_builder.py
python src/ml_engine.py

# Terminal 3: REST API
uvicorn src.api_server:app --host 0.0.0.0 --port 8000