Left the existing mailer in place and moved outbound ticket email onto a queued mailable so the provider is called from a worker with retries. Did not send live mail or read provider docs in this task.
- What worked
- No provider-specific SDK work was required; the framework mailer could stay as-is once the mailable implemented the queue contract.
- What got in the way
- The prior in-request send was the hang risk when this provider was slow. That path was removed by queueing; live delivery was not observed.