How Does Speech Detection Work?
March 16, 2011
Interactive voice response systems can be built and programmed to fulfill a multitude of functionalities. They can be programmed to integrate with a PBX (News - Alert) system, enabling them to perform call routing and auto-attendant functions whereby they process and parse incoming calls and transfer them to the correct destination. Additionally, companies frequently utilize IVR systems to place outbound calls to clients and consumers.
In situations where IVR systems are making outbound calls, they can be programmed, through the use and application of automatic speech recognition engines, to automatically detect what sort of telephonic contact they are making. Depending on whether they reach a live human, an answering machine or another auto attendant, programmers can invoke different types of scripts and program specific system actions.
Speech detection is put in place to detect callee type. Speech recognition engines are equipped to determine whether a human or a machine has received and been responsive to a call, and respond accordingly. Depending on the perceived end user, the system will employ branching logic to carry out a series of specific actions tailored to the type of response the phone call receives. The program can also be programmed to repeatedly place outbound automated calls at a specified interval until a live human is reached.
IVR systems utilizing speech detection are typically able to recognize whether they have reached a human or a machine within 2 to 3 seconds of when the call is taken. Developers might be hesitant to utilize speech detection in their application because of this short pause at the beginning of the call. Particularly companies utilizing interactive voice response systems for surveys would be hesitant to enable speech detection on automated outbound calls, as it might prompt hang ups, resulting in large non-response rates, and thus invalidating their supposedly random sample population.
Businesses of all types can utilize speech detection for various reasons. Market research firms can utilize detection capabilities when dispatching telephonic surveys. Systems could be programmed to place a series of calls until a live human was reached, or even leave a message with information on how to access the IVR survey at a later date. Medical professionals could utilize this system to dispatch outbound calls reminding patients that they were due for or had scheduled an appointment. Cable and phone companies could use speech detection capabilities to inform customers when there are problems with their service or additional service features available for purchase or upgrade.
In many instances, speech detection for outbound calls can enhance IVR system functionality by enabling the programming and incorporating of varying call responses based on situational occurrences. By eliminating enabling identification functionality, companies can free up workers for calls that absolutely require human interaction. Calls won’t be dropped or missed because the system is programmed to catalogue and document every type of situation they encounter, and they can call back as necessary.
TMCnet publishes expert commentary on various telecommunications, IT, call center, CRM and other technology-related topics. Are you an expert in one of these fields, and interested in having your perspective published on a site that gets several million unique visitors each month? Get in touch.
Edited by Juliana Kenny