We have a team of over 100 transcribers and there was no need for us to use voice-recognition, the caller appeared to be very pleased with our reply and related his experience using voice recognition, he tried using it for an audio recording of a focus group in which multiple speakers were participants, the length of the audio recording was 187 minutes
The caller said that the voice-recognition was able to document the first speaker, though not very accurately, but as soon as the second speaker started speaking, he could only see garbled text, he thought minor adjustments in the application were needed to make it work, but after having done that and re-starting the process the voice-recognition failed to recognize the first speaker and the garbled text continued.
It was disgusting, said the caller, after having spent $ 180 for the software and wasting time to tame the beast, the end result would be so disastrous. It was the least that he expected, and was sorely disappointed with its performance. If we were using voice recognition, he would not be interested is what he said to our customer executive. After we assured him that only human transcribers will document his audio, he went ahead and sent us the audio recording.