Nuance is launching two new Conversational AI Services, providing API-based access to constrained speech recognition and neural text-to-speech capabilities. In this article, we look at how adding these services gives our customers even more options for building AI-powered customer engagement applications in the cloud.
As pioneers in conversational AI for more than 25 years, we’re proud of our heritage. But we’re also committed to constantly redefining the state of the art by evolving our products, adding groundbreaking capabilities and enhancements, and making them easier to access and use.
Nuance Recognizer as a Service
The Nuance Recognizer engine has set the industry standard for automatic speech recognition for decades. By making it available as an API, we’re helping our customers move to the cloud and giving them the ability to quickly build and launch voice applications that outperform traditional speech-to-text (STT) solutions. It will also help our customers who have spent time creating and tuning their grammars protect those investments when they move to the cloud.
Nuance Recognizer as a Service offers more accuracy and better performance than alternative speech recognition solutions. For example, it excels at alphanumeric recognition (“Please say your 12-digit delivery tracking number.”) and it performs DTMF tone recognition, enabling organizations to contain more routine calls in the IVR. It’s an ideal AI backbone for IVRs that use menu options or directed dialog, or for small applications that don’t require lots of training to function effectively, as well as a convenient fallback option for first time callers or callers in noisy situations.
Nuance Neural Text-to-Speech as a Service
Being part of the Microsoft family allows Nuance to offer organizations trailblazing conversational AI capabilities developed in partnership with Microsoft’s AI research teams. Nuance Neural Text-to-Speech as a Service (Neural TTS) offers API-based access to next-generation speech synthesis from Azure Cognitive Services.
Unlike legacy text-to speech (TTS) engines, Neural TTS generates speech using advanced neural networks from the start of each conversation. That removes the choppiness associated with traditional TTS, resulting in synthetic voices that sound more human, reducing listening fatigue and increasing customer satisfaction. The Neural TTS system can also add a range of emotions to the responses it generates—from excited to sympathetic—helping customers feel understood and valued in automated conversations.
Neural TTS requires far less training and tuning than traditional TTS systems, and it can be trained in English and then learn to respond in other languages. That further reduces the time, cost, and effort of creating advanced conversational AI applications, and it enables our customers to have a single voice represent their brand globally.
Advanced conversational AI—and investment protection—in the cloud
These two new services are part of the Nuance Conversational AI Services portfolio, which gives organizations simple ways to create powerful new customer engagements across all channels—from intelligent voicebots in IVRs to conversational chatbots in digital interactions.
By accessing these services as APIs, our customers can also move existing applications to the cloud and take advantage of high levels of availability, scalability, and security. And because our Conversational AI Services integrate with leading IVR and Contact Center as a Service (CCaaS) platforms, organizations that have invested in building, tuning, and optimizing applications can move to the cloud knowing their existing solutions will continue to generate ROI.
Continuous evolution of conversational AI
Nuance continues to invest in evolving the Conversational AI Services portfolio by adding new services and optimizing the underlying algorithms based on our latest research. So, keep an eye out for more exciting announcements in the future as we discover new ways to power intelligent omnichannel customer engagement.