Sentry Retell — Miss Dawn QSC Dispatch

Retell AI conversation flow and agent configuration for Miss Dawn, the QSC AI dispatcher.

Overview

Miss Dawn is a voice AI dispatcher for Quality Service Company (QSC) field technicians. She handles:

• Post-job debriefs — 8-question structured debrief with silent coaching flags
• Help desk lookups — pricebook, schedule, customer, equipment, invoice, estimate
• Escalations — flags issues for office follow-up via ServiceTitan tasks

Architecture

Retell AI (Miss Dawn agent)
  └── Conversation Flow (5 nodes)
        ├── Greeting → identify tech by caller ID
        ├── Help Desk → 7 lookup tools
        ├── Job Debrief → 8 questions + save + escalate
        ├── Escalate → flag for office
        └── End Call
  └── 10 Custom Tools → Taylor AI Worker (Cloudflare)
        ├── /api/quinn/identify-tech
        ├── /api/quinn/appointments
        ├── /api/quinn/job
        ├── /api/quinn/customer-search
        ├── /api/quinn/pricebook
        ├── /api/quinn/equipment
        ├── /api/quinn/invoice
        ├── /api/quinn/estimate
        ├── /api/quinn/save-debrief
        └── /api/quinn/escalate

Files

File	Description
agent.json	Retell agent settings (voice, model, keywords, post-call analysis)
conversation-flow.json	Full conversation flow with nodes, edges, tools, and prompts

Agent Details

• Agent ID: agent_042204c3bac2eecc1fa37d2ffa
• Flow ID: conversation_flow_e6fe1560f922
• Voice: 11labs-Kathrine (en-US)
• Model: GPT-4.1-mini (cascading)
• Max call: 8 minutes
• Webhook: taylor-ai.lpeluso.workers.dev/api/quinn/webhook

Deployment

Agent is managed via Retell AI dashboard and API. Backend endpoints run on the Taylor AI Cloudflare Worker.

Status

• Agent + flow: configured
• Backend endpoints: deployed
• Phone number: not yet assigned
• Published: no