AI Agents · A home & moving-services company

Multi-Agent AI Voice Receptionist (Retell)

A multi-agent AI voice receptionist built on Retell for a home & moving-services company — it answers every inbound call, identifies intent, and warm-transfers to the right specialist agent or live team.

Retell AI — recruiting screening agent state machine

The problem

The company couldn't answer every call, and callers wanted very different things — quotes, existing-job questions, complaints, job applications, the owner. Missed and mis-routed calls meant lost jobs and frustrated customers.

Approach

A front-desk agent ("Scarlet") on a conversation-flow graph classifies caller intent from a welcome node and routes to specialist AI sub-agents, each with a transfer-failed fallback. A separate state-machine agent handles outbound recruiting screening.

What I built

• Intent routing from a welcome node into many branches — new quote, existing quote, booked move, customer service, HR/applicant, spam, FAQ, speak-with-owner
• Warm transfers to specialist sub-agents (lead intake, CS triage, item-claim, billing, executive assistant) with fallback paths if a transfer fails
• Knowledge-base-backed FAQ answering and structured post-call feedback capture
• Timezone-aware scheduling and rescheduling via a booking tool
• A separate outbound recruiting agent (state machine) that screens candidates and books interviews

Result

Shipped a voice front desk that triages every inbound call to the right place — specialist agent or human — so calls stop slipping through, with knowledge-base answers and clean escalation built in.