NVIDIA Riva sets a new bar for fully customizable voice AI
Whether for virtual assistants, transcriptions, or contact centers, voice AI services turn words and conversations into bits and bytes of business magic.
At GTC this week, NVIDIA announced new additions to NVIDIA Rivaa GPU-accelerated software development kit for building and deploying voice AI applications.
Riva’s pre-trained models are now offered in seven languages, including French and Hindi. Additional languages on the horizon: Arabic, Italian, Japanese, Korean, and Portuguese. Riva also brings improvements in the accuracy of English, German, Mandarin, Russian and Spanish. Additionally, it adds features like word-level confidence scores and speaker diarization – the process of identifying speakers in audio streams.
Riva is designed to be fully customizable at every stage of the voice AI pipeline to help solve unique problems efficiently. Developers can also deploy it where they want their data to be: on-premises, for hybrid multiclouds, at the edge, or in embedded devices. It is used by companies to boost services, efficiency and competitive advantage.
While AI for voice services has been in high demand, developer tools have lagged. More and more people are working and learning from home, shopping online and seeking remote customer support, which is straining call centers and pushing voice applications to their limits. Customer service wait times have recently tripled as staffing shortages have hit call centers hard, according to a 2022 Bloomberg report.
Advances in voice AI offer the way forward. NVIDIA Riva enables enterprises to explore broader deep learning models and develop more nuanced voice systems. Riva-based Speech AI apps offer an accelerated path to better services, promising improved customer experiences and engagement.
Growing demand for voice AI applications
The global contact center software market reached approximately $27 billion in 2021, a figure that is expected to nearly triple to $79 billion by 2029, according to Fortune Business Insights.
This increase is driven by the benefits that custom voice applications bring to businesses of all sizes, in nearly every industry – from global enterprises to original equipment manufacturers providing voice AI-based systems and cloud services, to system integrators and independent software vendors.
Riva SDK Accelerates AI Workflows
NVIDIA Riva includes pre-trained language models that can be used as-is or refined using transfer learning from the NVIDIA TAO Toolkit, which enables custom datasets in a no-code environment. Riva’s automated speech recognition (ASR) and text-to-speech (TTS) models can be optimized, exported and deployed as voice services.
Voice AI is making its way into more and more types of applications, such as virtual assistants and customer support chatbots, video conferencing systems, drive-thru food ordering, telephone retail, media and entertainment. Global organizations have adopted Riva to power voice AI efforts, including T-Mobile, Deloitte, HPE, Interactions, 1-800-Flowers.com, Quantiphi, and Kore.ai.
- T-Mobile adopted Riva for its T-Mobile Expert Assist — a personalized call center app that uses AI to transcribe real-time customer conversations and recommend solutions — for 17,000 customer service agents. T-Mobile plans to roll out Riva worldwide soon.
- Hewlett Packard Enterprise offers HPE ProLiant servers that include NVIDIA GPUs and NVIDIA Riva software in a system capable of developing and running complex voice AI and natural language processing workloads that can easily turn audio into information. HPE ProLiant and NVIDIA Riva Systems form a complete, world-class solution for running financial services and other industrial applications.
“To deliver the capabilities of NVIDIA Riva, HPE offers a Kubernetes-based NLP reference architecture powered by HPE Ezmeral software,” said Scott Ramsay, vice president of HPE GreenLake Solutions at HPE. “Delivered through the HPE GreenLake cloud platform, this system enables developers to accelerate the development and deployment of next-generation voice AI applications.”
- Deloitte supports customers looking to deploy ASR and TTS use cases, such as for order taking systems in some of the largest quick order restaurants in the world. It is also developing chatbot services for healthcare providers that will enable accurate and efficient transcriptions of patient questions and chat summaries.
“Advances in natural language processing enable the design of cost-effective experiences that enable focused, simple and natural customer conversations,” said Christine Ahn, principal at Deloitte US. “Our customers are looking for a streamlined path to deploying conversational AI, and NVIDIA Riva supports that path.”
- interactions has integrated Riva with its Curo software platform to create seamless, personalized engagements for clients across a wide range of industries, including telecommunications, as well as companies such as 1-800-Flowers.comwhich has deployed a voice AI order taking system.
- Kore.ai integrates Riva into its contact center as a SmartAssist Speech AI service, which powers its BankAssist, HealthAssist, AgentAssist, HR Assist and IT Assist products. Proof of concepts with NVIDIA Riva are in progress.
- Quantiphi is a solution delivery partner that develops captioning solutions using Riva for media and entertainment customers, including Fox News. He is also developing digital avatars with Riva for telecommunications and other industries.
Complex voice AI pipelines, simpler solutions
Voice AI pipelines can be complex and require coordination across multiple departments. Microservices must work at scale with ASR models, natural language understanding, TTS, and domain-specific applications. NVIDIA GPUs are ideal for accelerating these types of specialized tasks.
Riva offers software libraries for building voice AI applications and includes GPU-optimized services for ASR and TTS that use the latest deep learning models. Developers can merge these multiple voice AI skills into their applications.
Developers can easily access Riva and pretrained models via NVIDIA NGCa hub for GPU-optimized AI software, Jupyter Notebook models and samples.
Support for Riva is available via NVIDIA AI Enterprise, a cloud-native suite of AI and data analytics software optimized to empower any organization to use AI. It’s certified to deploy anywhere – from enterprise data center to public cloud – and includes global enterprise support to keep AI projects on track.
Try NVIDIA Riva with guided workshops on out-of-the-box infrastructure in NVIDIA LaunchPad.