After all the discussion over the past few days about SpinVox as a company, we thought it would be a good time to talk more abot our technology and how it works. As one of our investors, Julie Meyer, pointed out only yesterday, we’re a breakthrough company…with a strong IP <http://juliemariemeyer.typepad.com/in_the_game_julie_meyer/2009/07/swinging-for-the-fences.html> , and as we’ve grown, we’ve tried to walk the line between getting people to understand that we believe we are on to something special with our technology, whilst also trying to maintain a level of protection for our IP and systems.

However, it seems that the interest we’ve built up in SpinVox over the past few years has now grown to such an extent that there is a real desire to know in detail about what’s under the bonnet.  So today, we’ve released a lot more information about our technology and our approach to creating the SpinVox VMCS – Voice Message Conversion System – and the amazing performance improvements it has delivered over the past couple of years.

So to begin with, did you know that VMCS already knows more than 99 per cent of anything a user is likely to say in English or Spanish? It contains over two billion words and phrases derived from the equivalent of 72 years of audio training – making it the world’s biggest body of spoken language. This knowledge system is growing rapidly and accelerating our ability to automate and improve accuracy levels. It is now so advanced that VCMS only requires two per cent of the Quality Control agent input it required two years ago. This means that to support millions of users we’ve reduced the system’s need down to just a few hundred per market, as our automation has taken over 98% of the task.

We’ll be talking more about the technology that powers SpinvVox over the next few days, so please keep coming back to read more.

Daniel Doulton, co-founder