Veritone, an ambitious startup that is trying to do for online audio what Google did for Web search, said Nov. 17 that it is implementing Microsoft's advanced natural language processing engine inside its own cognitive media platform to provide state-of-the-art speech recognition to Veritone's customers.
Speech recognition is done well by only a few companies, with Nuance (Dragon), the largest provider, and Microsoft among them. It's very difficult to do it well; in fact, the industry is still researching ways for speech-recognition software to work with multiple voices. As of now, all of them only work with a single user, learning and remembering speech patterns and styles in order to produce accurate transcriptions.
The collaboration of Microsoft and Veritone software also will give Microsoft a broader distribution for its Azure Media Services. In addition to Indexer, Veritone's set of cognitive engines includes NLP, facial recognition, sentiment extraction, tonal analysis and fingerprinting.
The Newport Beach, Calif., startup launched itself last September and is providing an open ecosystem for developers, businesses and content owners to collaborate and monetize the insights gained by programmatic cognitive analysis of audio and video streams.
Veritone has a rather daunting mission: It wants to record, store and analyze all the spoken words in all public media. That's right; this means everything spoken in television, radio, online video, podcasts—you name it.
How can this be possible? Actually, while staying under the radar until now, the young company has been developing this functionality for two years and already has customers using it. In fact, Veritone processed its 1 millionth hour of media earlier this month. That, Veritone said, would be roughly equal to streaming YouTube non-stop, 24 hours a day for 110 years.
Veritone layers all the voice and video data it collects inside a temporal database for search and discovery, then wires it to what is essentially an open-ended action platform, or content management system (CMS), CEO and founder Chad Steelberg told eWEEK.
Veritone Media, the advertising arm of the company, is doing the first monetization of this technology for Veritone. Veritone Media delivers ads to the right listener/viewer by producing exact quantifications on the very inexact science of how media affects viewers/listeners/buyers and persuades them to do or buy something.
"In the case of Veritone Media, we deliver ads. In the case of other platforms, we use Web hosts to talk to CRM apps; in education, we notify students in the case of the education platform, and so on," Steelberg said.
Veritone expects to process millions of hours of audio and video content by collecting and processing content in the Microsoft Azure cloud. By integrating within the CMP, Indexer expands its reach into a vast ecosystem of media to process, generating significant revenue streams for Media Services. As more media is processed by CMP, the more accurate the returned results are. As the breadth of the data and the IQ of the CMP increases, so does the benefit to its customers, Veritone said.
The net result is actionable intelligence in near real time that can directly impact Veritone's CMP customers' bottom line.