Since this is the first post to our new Vlingo blog, I figured I should use it to give all of you some info on the background and history of Vlingo.

While the company is less than three years old, the story actually goes back a lot farther than that. A lot of the people at Vlingo have worked on speech recognition for many years – either as researchers or building the technology and products which have been used for call centers or built into cellphones and cars. In the late 1990’s as there were starting to be reasonably capable PDAs and the beginnings of wireless networks, a lot of us realized that eventually mobile devices would be the place this technology would really become useful. But, of course, it was way too early! Thinking about some of the things we were talking about back then, we had the overall concept right, but the only way we could actually build something was on some hefty PDA with some extra hardware just for the speech recognition part.

In 2006, things were looking much better for actually deploying interesting applications on mainstream mobile devices making use of network-side services and content (of course, this isn’t completely true even today – there are still constraints on what can be deployed on mobile devices and wireless data network, while mostly pretty good, are still not as reliable as we would all like). John Nguyen (who had worked with me at SpeechWorks) and I decided that the time was indeed right, and that a startup would be the right place to develop the next generation of speech technology for mobile devices. We formed the company, hired Han Shu (from MIT) and Joe Cerra (from Tufts), got our first round of financing in June 2006 and then started working away on our first prototypes.

While we could have taken the approach of making incremental improvements on the existing sorts of mobile speech interfaces (things like voice dialing, and other specific speech-enabled applications), we decided we needed a much bigger goal – to fully enable mobile devices and applications with a new speech-enabled user interface. It’s been our view that user interface is quickly becoming the only remaining barrier to what applications and services can be made available on mobile devices, and that speech input and output can play a significant role in solving this user interface challenge.

But, to do this in a way which works across all applications, we had to make big changes to the core technology. Until our first prototypes, it was widely believed that the only way you could make speech recognition work well enough in mobile environments was to constrain the speech recognizer as much as possible and then construct the application around those constraints (“please say the city name now”). While this can work with individual applications as long as the user plays along, it just doesn’t scale across all the things we want to do – and users don’t like to play along!!

So, most of our initial work at Vlingo was in developing the core technologies which allowed us to get rid of these constraints. Once we did this, a whole range of applications became possible for us. These included our first commercial deployment with Yahoo! in April 2008 (voice enabled open web search in the Yahoo! oneSearch product) and then Vlingo for Blackberry in June 2008 and Vlingo for iPhone in December 2008.

We now have a great team of close to 50 people working away on improving the core technology, adding languages, adding functionality to the product, and porting to an increasing number of phones. We think what we have deployed so far is just the beginning of what is possible and look forward to releasing a continuing series of products and seeing how they are adopted in the market.

We hope this new blog will be a good way to provide more information on what we are doing and most importantly to get feedback from others. So, please take a look here from time to time and let us know what you think!

Mike Phillips – Co-Founder and CTO