Verbumware Inc. is a start-up company that specializes in up-scaling audio applications to the mass market.
We offer a fast and efficient hardware-accelerated cloud-based speech recognition service that can be customized to suite the needs of a specific application by a proper choice of acoustic and language models.
Our technology is based on application of the massively parallel computing architecture of modern Graphical Processing Units (GPUs) to the problem of speech recognition and interpretation. Our system does recognition orders of magnitude faster than people are able to speak.
Our speech recognition platform can use models developed with publicly available open-source toolkits. Thus, our customers can have a complete control over the intrinsics of the recognition system they use.
As experts in the field we also offer assistance in development of the statistical (acoustic, language, understanding) models for speech recognition. We also provide consultancy in application development and technology integration.