Skip to main content

Analysis Object Model

Core domain entities and relationships, one diagram per backend service. The three Spring services were modeled with a UML tool; py-intelligence is expressed as Mermaid since it's generated directly from the current code.

spring-logbook​

Owns the system_logs table: consumes ingestion events off RabbitMQ, persists them, and serves the deployment history timeline.

spring-logbook class diagram

spring-alerts​

Owns the incident_status table. Uses a Strategy pattern (AlertStrategy implementations like MetricThresholdStrategy) so new alert conditions can be added without touching the core rules engine.

spring-alerts class diagram

spring-ingestion​

The synchronous entry point: validates incoming payloads, assigns the correlation logId, and publishes to RabbitMQ. Holds no persistent state of its own.

spring-ingestion class diagram

py-intelligence​

The GenAI service. Intelligence orchestrates prompt building, model selection, and response parsing; Model wraps each LLM backend (local Qwen via llama.cpp, Gemini, OpenAI-compatible fallback); DB wraps MongoDB access for both RAG documents and persisted analyses.