Sign up

Documentation

AI Routing

AI routing uses machine learning to optimize outcomes over time, adapting to changing conditions in provider networks and markets.

What you specify

Optimization goal

  • Maximize delivery rate
  • Minimize cost per delivered message
  • Minimize latency
  • Maximize engagement (read receipts / replies)

Constraints

  • Budget caps
  • Provider exclusions
  • Destination restrictions
  • Channel preferences

How it works

  • 1Telemetry collected from all messages routed through TrustRouter
  • 2Continuous A/B testing across routes (with configurable exploration rates)
  • 3Model retraining on aggregated performance data
  • 4Prediction of delivery probability, latency, and engagement per route

Transparency

Every AI routing decision is logged and auditable. You can inspect why a route was selected and override the model's decision with manual or policy routing at any time.