THERE ARE TWO AREAS 
                in a company which an executive can examine for potential benefits 
                from voice technology. The first area is daily office routine. 
                A second area is transactions with customers.
               Daily office routine:
               Think about the kinds of commands you could give to a computer 
                that represent things that you really need to do every day. Here 
                are some examples: "What's the price of my IBM stock?" "get my 
                Internet mail" "what's the status of the Hong Kong order" "what's 
                the status of the Lambert case?"
               Computers can be trained to listen for these kinds of commands 
                and deliver the expected results. Off-the-shelf products already 
                will take dictation and transcribe them into fully formatted letters 
                without special training.
               Customer transactions:
               Many businesses have significant interaction with customers 
                via the telephone. A simple study can often reveal where speech 
                recognition and speech synthesis approaches can have major bottom 
                line impact on these transactions and yet keep customer satisfaction 
                higher than the old annoying "press 1 for service" messages. 
                Generally, there are no off-the-shelf solutions, but there are 
                off-the-shelf components which can be assembled into useful solutions. 
                Solutions in this category have been historically called telephony. 
                For a good example of how well this could be done for your company, 
                call this interactive computer demonstration number: (610)-648-2951 
                (demonstration provided courtesy of UNISYS). 
              Company web sites can now also be provided with a high degree 
                of engaging speech assisted interaction.
               About accuracy:
               Speech recognition has two components: accuracy and fluency. 
                Accuracy is the aspect of a computer system which allows it to 
                correctly identify each word you speak as a word it knows. Fluency 
                is the aspect concerning the range of grammar and vocabulary which 
                the computer system knows. Both aspects are important, but as 
                an executive you'll be most comfortable with systems which have 
                high fluency.
               Why? High fluency places fewer demands to memorize special command-and-control 
                phrases. That means lower training costs and therefore quicker 
                productivity boost.
              Off-the-shelf dictation products have remarkable accuracy but 
                each vendor incorporates different levels of fluency. Dragon's 
                NaturallySpeaking currently has the most fluency for command-and-control. 
                The accuracy and fluency for your own telephony applications can 
                be made as high as needed in proportion to the return on investment 
                required.