Changes / View Change
- Created:
- Status: staging
- Faster delivery: scheduled-to-sent latency drops from tens of seconds to 1-3 seconds in typical cases.
- Tighter tail latency: bulk blasts that previously took minutes to drain now complete in seconds.
- Higher throughput ceiling: agencies sending thousands of messages per minute are no longer capped by worker capacity.
- Improved recovery: stuck messages are detected and retried automatically every 5 minutes.