How do startups need IaaS: Speech Recognition Services
This is one of our practical rubrics, in which we share examples of many years of cooperation with Russian and foreign companies. For example, earlier we talked about cases Hotels.ru and PickPoint .
Today we would like to introduce you to the Speech Technology Center (MDG) , which you should have already met on Habré - they have an excellent blog.
/ photo Per-Olof Forsberg CC
')
The MDGs were founded by acoustics specialists who have invested their scientific potential and development in the company. The work of the MDG team proved to be in demand at the international level - the characteristics of the products were not inferior, and even exceeded the capabilities of foreign analogues.
Today's MDGs are a company with a history of more than 20 years and a leading developer of high-quality recording, processing and analysis of audio-video information, speech synthesis and recognition. The company's products are based on IT-GRAD's IT infrastructure, which makes it possible to quickly develop new products and use a pool of computing resources with high availability.
Now look at the projects themselves.
This is a service that allows you to voice any textual information with a synthesized voice. It works on the basis of cloud capabilities and even has its own API .
Now this Internet service can work with such complex structures as homographs . He understands timbre and tone using SSML tags and recognize emotional responses.
Voices, voiced text understand even the designation ":)" and ": (" - they laugh and sigh. Implemented support for several languages. These include, for example, Kazakh.
VoiceNavigator project - software for IVR contact center systems. Able to recognize and synthesize Russian speech (translates speech into text and vice versa).
In simple words, instead of typing the desired number on the keyboard (“to learn about our services, press 1”) you can control the context menu with voice commands. The recognition algorithm reads the request subject and synthesizes the response. The technology is also well suited for queuing systems.
Within the framework of cooperation with MegaFon, the MDG company proposed a solution for the customer service service - a virtual consultant. Elena (so decided to call the service) can process more than a million inquiries per year and save the company's customers thousands of hours that they could spend waiting for a response from a living person.
At the moment, Elena is a pilot project based on MDG technologies.
Today we would like to introduce you to the Speech Technology Center (MDG) , which you should have already met on Habré - they have an excellent blog.
/ photo Per-Olof Forsberg CC')
What is "speech"
The MDGs were founded by acoustics specialists who have invested their scientific potential and development in the company. The work of the MDG team proved to be in demand at the international level - the characteristics of the products were not inferior, and even exceeded the capabilities of foreign analogues.
Today's MDGs are a company with a history of more than 20 years and a leading developer of high-quality recording, processing and analysis of audio-video information, speech synthesis and recognition. The company's products are based on IT-GRAD's IT infrastructure, which makes it possible to quickly develop new products and use a pool of computing resources with high availability.
Since the MDGs regularly conduct high-throughput testing of products, as well as implement external projects, the choice in favor of the cloud is completely justified.
- MDG team
Now look at the projects themselves.
Voicefabric
This is a service that allows you to voice any textual information with a synthesized voice. It works on the basis of cloud capabilities and even has its own API .
Now this Internet service can work with such complex structures as homographs . He understands timbre and tone using SSML tags and recognize emotional responses.
Voices, voiced text understand even the designation ":)" and ": (" - they laugh and sigh. Implemented support for several languages. These include, for example, Kazakh.
VoiceNavigator
VoiceNavigator project - software for IVR contact center systems. Able to recognize and synthesize Russian speech (translates speech into text and vice versa).
In simple words, instead of typing the desired number on the keyboard (“to learn about our services, press 1”) you can control the context menu with voice commands. The recognition algorithm reads the request subject and synthesizes the response. The technology is also well suited for queuing systems.
Virtual consultant of "Megaphone"
Within the framework of cooperation with MegaFon, the MDG company proposed a solution for the customer service service - a virtual consultant. Elena (so decided to call the service) can process more than a million inquiries per year and save the company's customers thousands of hours that they could spend waiting for a response from a living person.
At the moment, Elena is a pilot project based on MDG technologies.
A significant part of calls to the contact center is typical and can be automated with the help of modern speech technology.
Working in close cooperation with MegaFon, the MDGs created not just a guide with voice control, but a real “person”, communication with which is as close to natural as possible.
- MDG team
Source: https://habr.com/ru/post/268343/
All Articles