Voice Messaging

Outbound voice, AI voice agents, and voicemail drop — orchestrated by RME, recorded in Salesforce, and analyzed by Voice Intelligence.

Capabilities

• Text-to-speech outbound calls with SSML control
• AI voice agents that can read from and write to Salesforce during the call
• Voicemail drop with answering-machine detection
• Call recording, transcription, and sentiment via Voice Intelligence
• DTMF capture for confirmations and simple IVR flows

Placing a call

RME_Message.builder()
    .channel('VOICE')
    .toPhone(contact.Phone)
    .toContact(contact.Id)
    .voiceScript('appointment_reminder_v3')
    .parameters(new Map<String, String>{
        'firstName' => contact.FirstName,
        'when'      => '<say-as interpret-as="date">2026-08-06</say-as>'
    })
    .amd('VOICEMAIL_DROP')       // answering machine detection behavior
    .send();

Script definition

<script name="appointment_reminder_v3" locale="en-US" voice="neural-en-US-Aria">
  <say>Hello {{firstName}}, this is Acme Health confirming your appointment on {{when}}.</say>
  <gather digits="1" timeout="6">
    <say>Press 1 to confirm, or 2 to reschedule.</say>
  </gather>
  <on digit="1" event="CONFIRMED" />
  <on digit="2" event="RESCHEDULE_REQUESTED" />
  <fallback say="We will follow up by text message." channel="SMS" />
</script>

Events

initiated, ringing, answered, machine_detected, dtmf, completed, no_answer, busy, failed. Each event publishes RME_Delivery_Event__e, so a Flow can create a Task or update a Case the moment a call completes.

Compliance

Voice traffic is subject to calling-time restrictions, do-not-call lists, and recording consent. RME enforces quiet hours and DNC suppression before dialing, and stores the consent basis on the message record. See Compliance.

Continue to Push Notifications.

Last updated 1 month ago