System Health

Data source monitoring and status

Backend API
Data Sources
0 registered
Degraded
All healthy
Signal Suppression
Inactive

Data Sources

Execution Integrity
Spread Bucket
no snapshot
Feed Consistency
no data
Spread
-
Divergence Checks
0
Inconsistent Windows
0
Pipeline Architecture

Data flows through the candle-close event pipeline. Each candle close triggers all analysis layers (ICT, Order Flow, CNN, Strategy Rotation, COT) which produce calibrated probability signals. These are aggregated into a composite signal weighted by Brier-optimized layer weights.

When any data source is stale or degraded, the system suppresses signal generation to prevent unreliable outputs. The health monitor continuously checks source freshness against configured thresholds.

OANDA CandlesHMM RegimeCNN InferenceCOT DataCalibration Worker