First-generation VUIs

The big break came in 1984, when SpeechWorks and Nuance introduced interactive voice response (IVR) systems. IVR systems were able to recognize human voices over the telephone and carried out tasks given to them (Roberto Pieraccini and Lawrence Rabiner 2012, The Voice in the Machine: Building Computers That Understand Speech). You will be able to recognize IVR systems today when you call major companies for support. For example, when you call to make a hotel reservation, you will be familiar with "Press 1 or say reservation, Press 2 or say check reservation, Press 3 or say cancel reservation, Press # or say main menu." In the '90s, I remember working on my first VUIs in an IVR system. To develop the IVR system, I had to work with the Microsoft Speech API (SAPI), http://bit.ly/2osJpdM. With SAPI, I was able to perform text to speech (TTS), where the voice received from the user was translated into text in order to evaluate the user's intent; then, after evaluating the user's intent, a text message was created and converted back to the voice to relay the message to the user on the telephone.