Check out Marcello Gallucci's statement
CTO of Gennex
Check out Marcos Trestin's testimonial
CEO of Trestto Technology and Innovation
With the new feature of Speech Recognition – Sentiment Analysis – it is possible to
identify the customer's mood study in marketing, sales and collection actions.
Count on Speech Recognition – an extremely accurate online resource that recognizes speech in Portuguese and/or Spanish. To achieve a real approximation and greater user engagement, the systemtema was trained with the speech of thousands of Brazilians, withtemplanning the most different regionalizations and accents from all over Brazil.
Earn tempo, inclusion and customer satisfaction through speech recognition (STT/ASR) in automated service. Increase efficiency and reduce temtype of service with automatic and intuitive interactions. In addition, you promote the inclusion of people with little knowledge in technology or even the visually impaired.
Listen to your assistant.
Integrate digital media and provide accurate answers to customers.
Facilitate and expedite the service effectively.
Faster and more accurate to get important data.
Convenience and speed of service.
Free your hands. You ask, he obeys.
The solution for those who need to forward high volumes of messages.
Implement a solution that will save tempo of your users.
Accuracy and robustness in any environment
Whether in an isolated room or in an environment with a lot of sound interference, the systemtema is accurate – thanks to machine learning technology. No pre-processing of the signal or noise cancellation is required.
Flexible and supported integration
Speech Recognition easily integrates into any application or device via HTTP REST interface, MRCP V1 and V2 or Websocket. In addition, CPQD provides the SDK (devkit) for software development in C++ and Java languages.
Different usage scenarios
The product recognizes free speech for generic scenarios and without the need to create syntax rules and speech in grammatical mode, in which the grammars are written in the SRGS/SISR standard as a way of customization for a given context of use. When the suntema recognizes something that is not speech, it automatically removes the initial and final silence from the audio.
efficient processing
the suntema processes audio as it is received and returns partial results while speech is recognized. In addition to online recognition, the sistema returns text from pre-recorded audio recognition in linear PCM format, with a sampling rate of 8 kHz or 16 kHz.