RealSpeaker Inc. develops the audio-visual speech recognition application RealSpeaker. Our solution uses additional video information, which allows as to improve voice recognition accuracy by at least 20-30 per cent. This makes our patent pending technology way ahead of existing competitions. We have been working for 2.5 years. We have released beta version of our product for Windows OS, which is currently in use by about 3k paid users from Russia. In this year, we are going to reach over 35 paid users with a revenue over $1M by the means of English, Spanish and Deutsch speaking people too.