How It Works

We run controlled probe traffic and normalize telemetry into a consistent pipeline so comparisons remain meaningful over time.

Probe Execution: Distributed probe accounts send and receive controlled test messages across target providers.
Timestamp Capture: Event timestamps are captured at generation, submission, acceptance, and mailbox-observed delivery points.
Normalization: Raw events are transformed into a common schema so provider and region comparisons stay apples-to-apples.
Aggregation: Stage timings and outcomes are aggregated into rolling windows for trend detection and incident visibility.
Quality Controls: Duplicate suppression, clock sanity checks, and schema validation reduce noise and improve reliability.