SPEECH RECOGNITION SYSTEMS

Speech recognition software has replaced a lot of secretaries in many companies

By LAMIKA ZUBERI
Apr 24 - May 07, 2006

Commercial systems for speech recognition have been available since the 1990s. Point of Sale machines have been talking back change since early 1990's at major stores in USA and Europe. In these stores they made the task of the cashier much easier and at the same time shoppers would feel confident that they are getting the right change back. This helped improve the quality of customer interaction. Now voice is being used in call center software e.g. to choose the dial pad options, e.g. For sales press one, for customer services press two etc. Now you do not need to press one or two, you can just say it with voice and the call center software would recognize your voice as a command and forward the call to the respective department of your choice. Companies in USA had their first implementations in banks and telecom companies. But in Pakistan the speed at which companies take to modern technology is slow.

In the west this technology is being implemented everywhere. It has been around for more than 25 years, but why is it more popular now. This is so because now there is computer hardware available to suit this technology and also now it is more affordable. Previously computer hardware was very expensive, but with the developments in Central Processing Units and Digital Signaling Processors, it is affordable.

In the west, physicians may use speech recognition system to replace their transcription personnel or their transcription services. Using this technology, physicians gain a lot. First is cost. It is very expensive for transcription services, in the west. A lot of the medical transcription work used to be outsourced to India and some to Pakistan. Speech recognition systems seem to be a viable choice if they are 99.9% reliable and accurate.

Speech recognition software has replaced a lot of secretaries in many companies. Such softwares can type 160 words per minutes using your voice. Microphone is used as an input device instead of a key board, and the text is produced much faster due to the fact that speech is faster than dictation and typing. Modern technology helps organization save cost. In the developed countries, following speech recognition software are being popularly used in the respective field as under:

Accounting

Dragon Dictate

Corporate Entities

IBM Voicetype

Developer

Dragon Tools

Disabled

Kolvox OfficeTALK w/Dragon

Journalism

IBM VoiceType

Legal

Kolvox LawTALK

Medical

Kurzweil VoiceMED

Speech recognition can be used as a great presentation tool. When you are on your sales trip you may need to add a cool feature to your presentation. You can give your presentation file a name and call it; the system will open your file. People with disabilities like carpal tunnel syndrome and other disabilities can be as productive as normal people.

Call centers have a self serve menu, which is driven by the caller. With voice it would be even easier. Companies would not need to hire a lot of agents. The processes will be further automated and many callers would be able to drive their call all the way to close. With the increased sophistication of the technology, speech applications would be doing a lot more than they used to do. Many live agent calls and transactions can be easily automated, which means cost savings. Speech recognition provides the ability to recognize the customer's need faster. It enables us to do transactions that could not have been done before, such as being able to recognize what city you are from rather than going through a menu of all cities. It also enables you to do some problem solving that very few applications are capable of performing. There are still problem-solving issues that companies have with their customers that will require live agents. This will help companies save costs.

Voice self-service applications are consistent. They have a consistent feel, a consistent tone of voice, a consistent delivery of the brand. That consistency helps in terms of increased customer satisfaction. On the other hand, the interaction with a speech recognition application can also be frustrating because the recognition engine may perform at an 80-90 percent effective level. This means that 10-20 percent of the callers would be frustrated because the voice recognition engine would mess up their call.

If we look at voice over IP, we can see that companies are making investments in VoIP and getting it to their in-house call centers. There is a lot to be said for being able to work with a call center technology provider who not only sees best practices across a number of clients within a particular vertical market, but is also able to raise the level of customer interaction performance for clients because of that knowledge. The transfer of knowledge and expertise helps the call center technology provider's client improve client interaction which translates into better revenue and smooth operations.

In call center technology, the financial services industry has certainly been an early adopter. Large banks and financial services organizations have been leaders in technology investment over the last 10 years. They were one of the earliest industries to provide customer information and transactions online. Because of that early experience, financial services firms are uniquely positioned to transition to self-service through voice channels. Most recently, we are starting to see awareness from the vertical markets that touch large groups of consumers: telecoms, cable, satellite and ISPs are all moving fairly quickly to adopt call center solutions.