QOne Electronic Queue offers its own speech synthesis service to deploy a voice assistant in addition to customer flow automation systems. Our solution is a highly efficient caching service with the ability to automatically replenish the message database through external speech synthesis services (Google Text-to-Speech or Responsive Voice).
Advantages of a speech synthesis service
- Low cost of generating new files. For example, for a database of audio files based on Google Text-to-Speech, the price is 16 USD per 1 million characters
- Reducing the cost of service ownership by segmenting messages. Each unique segment is voiced only once
- Proprietary message normalization algorithms, which can further reduce calls to external services
Google Text-to-Speech
- Supports 120+ languages and language varieties
- Accurate pronunciation, formatting phrases depending on the language situation (e.g. dates, phone numbers)
- High level of speech imitation, close to real life
- Support for Speech Synthesis Markup Language (SSML) based on XML
Responsive Voice
- Sound analysis service from Responsive Voice company
- Support for 42 languages
- Text-to-speech conversion based on hidden Markov models
- Intonation is determined by a neural network trained on many real-life examples
Service based on pre-recorded texts
If necessary, our electronic queue can use a service based on pre-recorded texts.
This approach is used in companies with completely closed circuits. Relevant for banks and other organizations with high security requirements.
The service used in the QOne Electronic Queue system provides a simple and powerful API that allows you to receive mp3 voice messages generated based on a third-party voice platform.
Voice messages are transmitted in real-time or generated as a pre-created cache from audio files. The cache is built on proprietary algorithms and the service's internal database.
Modern voice services at a relatively high level solve the issue of simulating pronunciation close to natural - problems with recreating voice timbre, smoothness of sound, and intonation are eliminated. Our solution leverages all the advantages of modern technologies for recognition and vocalization from leading global developers.