Text-to-speech (TTS), or speech synthesis, is often used in IVR applications to provide callers with information that changes often or is dynamically generated from a database system. 

For instance, if a caller needs to access his or her bank account balance over the phone, a TTS engine can provide the account balance, which changes as funds enter and leave the account. 

Using the TTS engine to provide information through speech saves time and money while eliminating the need for programmers to concatenate recorded audio to provide ever-changing information.

In an IVR application or call flow, TTS typically works alongside recorded audio. Recorded audio prompts callers to enter information through their touchtone keypad or through speech recognition while providing them with instruction for navigating through a menu of options.

When selecting a text-to-speech engine, most IVR developers and businesses choose a voice that most closely resembles the voice and speech patterns of the individual who recorded the static audio prompts. This helps ensure a more user-friendly experience when the call flow transitions from recorded audio to TTS. 

The Plum IVR hosted service supports three different TTS engines with more than 40 voices to choose from so developers can pick a voice that closely resembles their recorded static audio prompts. 

The TTS engines running in the Plum hosted IVR service include Nuance, Cepstral (News - Alert) and AT&T natural voices.  Supported languages include English, French, German, Spanish, Portuguese and Mandarin Chinese.  Japanese and Korean will follow shortly.