Glossary
Text-to-speech
In short
Text-to-speech (TTS) is technology that turns written text into spoken audio. Modern AI voices sound close to a real person, with natural pacing and intonation.
In a voice agent, text-to-speech is the mouth: the AI decides what to say as text, and TTS speaks it to the caller. Speed matters as much as quality, because long pauses make any call feel robotic.
Businesses can usually choose the voice (accent, tone and pace) to suit their brand and their callers, and some voices can switch language in the middle of a call.
Older phone systems used stiff, obviously synthetic voices, which is why many people still expect automated calls to sound like robots. Current voices have largely closed that gap.
A worked example
A physiotherapy clinic picks a calm, warm voice for its phone agent, because many callers are in pain and a brisk voice feels wrong. The same clinic sets a slightly slower speaking pace for older patients. Both are text-to-speech settings, not scripts.
When it matters, and when it doesn't
Voice choice matters most when callers are anxious, older or calling about something personal. For quick, practical calls — opening hours, a delivery slot — speed and clarity matter more than the exact voice.
Common questions
Can I choose the voice?
Usually, yes — gender, accent, tone and pace — from a library of voices.
Can it use my own voice?
Some providers offer voice cloning from a short recording, with your consent. Many businesses prefer a neutral voice so callers are not misled into thinking they reached a specific person.
Does it speak other languages?
Yes. Many voices speak several languages, and some can switch mid-call.