Configure Voice, Recognition, Recording and Transcription

Choose voice and speech-recognition settings and manage call recording, transcripts, access and retention.

What this guide helps you do

Choose voice and speech-recognition settings and manage call recording, transcripts, access and retention.

Who should use it

Workspace owners, managers, automation users and authorized technical users responsible for AI configuration or customer workflows.

Before you begin

Use the correct workspace and minimum required role.
Use a business-controlled provider account where applicable.
Understand provider billing, usage and data settings.
Use fictional Pakistan test data.
Define human handover before enabling automatic replies or calls.

Main concepts and fields

Concept / Field	Meaning or Use
Voice Provider	Text-to-speech service.
Voice	Selected speaker.
Recognition Provider	Speech-to-text service.
Language	Expected caller language.
Recording	Stored call audio.
Transcription	Generated text.
Retention	How long call data is kept.

Step-by-step instructions

Choose the voice provider and voice.
Choose language/recognition behavior.
Test English, Urdu, Roman Urdu and mixed speech where required.
Enable recording only when authorized.
Add caller disclosure.
Enable transcription where required.
Restrict access.
Set retention and deletion process.
Test audio quality and transcript accuracy.

Test and verification

Test quiet and noisy calls.
Test interruption and silence.
Review recording access.
Review transcript accuracy and retention.

Common issues and solutions

Issue	What to Check
Poor recognition	Language, audio quality and provider.
Recording missing	Provider capability, setting, storage and permission.
Transcript inaccurate	Audio and recognition configuration.

Last reviewed: 2026-07-15