How deaththekid1614/netwroxia works
Netwroxia is an offline AI network-operations copilot for banking networks. Its main workflow collects network telemetry, predicts faults and time-to-impact, explains risk, and supports incident handling and remediation; operators inspect results in a Streamlit mission-control dashboard. The README also documents beacon readiness, banking-impact analysis, RCA, explainability, analytics, MLOps, and reporting. Some stage links are documented only at a high level; unsampled implementation details are represented by their listed source paths, without asserting unshown calls.
Main entry points: predict.py Fault prediction · orchestrator.py Remediation orchestration · incident_store.py Incident records · incident_engine.py Incident lifecycle
Components 25
Network inputs
- Bank network
- Telemetry store
- Beacon monitoring
beacons/beacon_collector.py - System readiness
beacons/system_check.py
Prediction and copilot
- Metric fetching
ml/data/fetch_metrics.py - Fault prediction
ml/inference/predict.py - Offline copilot
copilot/llm/inference.py - Knowledge ingestion
copilot/rag/ingest_documents.py
Incident intelligence
- Health scoring
analytics/health_score.py - Banking impact
impact/impact_analyzer.py - Root-cause analysis
rca/rca_output.py - Risk explanation
explain/explanation_output.py
Incident response
- Incident lifecycle
incidents/incident_engine.py - Incident records
incidents/incident_store.py - Remediation orchestration
remediation/engine/orchestrator.py - Action executor
remediation/engine/executor.py - Approval and rollback
remediation/guardrails/approval_gate.py - Audit trail
audit/audit_logger.py
Operator and analytics
- KPI and SLA analytics
analytics/sla_tracker.py - Model monitoring
mlops/performance_tracker.py - Operational reports
reports/incident_report.py - Mission-control UI
dashboard/app.py - Dashboard pipeline runner
dashboard/utils/pipeline_runner.py - Dashboard telemetry client
dashboard/utils/influx_client.py
Other components
- NOC engineer
Connections 27
- Bank network → Telemetry store: exports metrics
- Telemetry store → Metric fetching: provides metrics
- Metric fetching → Fault prediction: supplies features
ml/inference/predict.py - Fault prediction → Offline copilot: feeds predictions
- Knowledge ingestion → Offline copilot: supplies knowledge
- Fault prediction → Health scoring: informs scoring
- Fault prediction → Banking impact: informs impact
- Fault prediction → Root-cause analysis: supports RCA
- Root-cause analysis → Risk explanation: explains risk
- Root-cause analysis → Incident lifecycle: creates incidents
incidents/incident_engine.py - Incident lifecycle → Incident records: updates records
incidents/incident_engine.py - Incident lifecycle → Audit trail: logs changes
incidents/incident_engine.py - Remediation orchestration → Incident lifecycle: runs lifecycle
remediation/engine/orchestrator.py - Remediation orchestration → Action executor: invokes executor
remediation/engine/orchestrator.py - Remediation orchestration → Approval and rollback: checks approval
remediation/engine/orchestrator.py - Remediation orchestration → Audit trail: records cycle
remediation/engine/orchestrator.py - Incident records → KPI and SLA analytics: supports trends
analytics/trend_engine.py - Incident records → Model monitoring: provides incidents
mlops/performance_tracker.py - Incident records → Operational reports: provides records
reports/incident_report.py - Audit trail → Operational reports: provides events
reports/incident_report.py - Approval and rollback → Operational reports: provides snapshots
reports/incident_report.py - NOC engineer → Mission-control UI: views operations
- Mission-control UI → Dashboard pipeline runner: runs pipeline
dashboard/app.py - Mission-control UI → Dashboard telemetry client: queries telemetry
dashboard/app.py - Dashboard pipeline runner → Fault prediction: runs predictions
- Beacon monitoring → Bank network: monitors routers
- System readiness → Bank network: checks readiness