Send failed production test alerts to on-call engineers before SLAs lapse

Your production pipeline posts failed jobs into noisy feeds, causing delayed triage and missed fixes. On-call engineers get owner-tagged alerts so incidents are acknowledged before SLAs lapse.