AI Modules
Overview of the AI and machine learning modules in the BT Management System.
Available Modules
- ML Models — Training and inference pipeline for predictive analytics
- Anomaly Detection — Real-time anomaly detection on device telemetry
- Behavioral Analytics — User and device behavior pattern analysis
- Smart Maintenance — Predictive maintenance recommendations
Architecture
The AI modules are implemented as NestJS services within the server-nest project, using the following Prisma models:
| Model | Purpose |
|---|---|
MLModel | ML model definitions and configurations |
TrainedModel | Trained model artifacts and metadata |
MlPrediction | Prediction results and history |
AnomalyDetection | Anomaly detection rule configurations |
Anomaly | Detected anomaly records |
DetectedAnomaly | Anomaly detection results |
DetectedPattern | Identified behavioral patterns |
BehavioralAnalytics | Behavioral analysis results |
SecurityAnalysis | AI-driven security analysis |
SmartMaintenance | Maintenance prediction configurations |
MaintenancePrediction | Maintenance prediction results |
Multi-Tenancy
All AI models and predictions are automatically scoped to the authenticated user via Prisma middleware. Each tenant's data is completely isolated.