UBC — AI Solutions

Audit — Dynamic Documentation Platform

Living documentation that stays in sync with engineering, cloud, AI, and compliance activity. Documents update themselves as the organization changes.

Back to library
Runbook
Auto-generated
Fresh

Runbook — Sarah Agent Outage

Detection, triage, and recovery for Sarah voice agent unavailability.

Owner: SREApprover: Platform EngVersion: auto-3.0Updated: 12/28/1969

Detection

CloudWatch alarm sarah-conversation-init-5xx > 1% for 5m.

Triage

  1. Check ElevenLabs status page
  2. Check OpenAI Realtime status
  3. Roll back to previous prompt version

Recovery

Enable fallback IVR routing via Twilio flow sarah-fallback.

Suggested by AI
No suggestions right now.
Frameworks
SOC 2