Inference Launches Inference Studio 6.5

Inference Solutions, a provider of intelligent virtual agents (IVAs), today released version 6.5 of Inference Studio, which includes new features for scaling virtual agent applications across voice and digital touchpoints.

With Studio 6.5, businesses can use a single code-free platform to develop and deploy IVA applications for voice, SMS, WhatsApp and, now, web-based chatbots.

Studio also includes a Task Library of templates that businesses can use to develop IVA applications, such as order lookups, credit card payments, or frequently asked questions. Applications can also be built from scratch using drag-and-drop tools, and service providers can publish their own templates to the Task Library. With Studio, users can create one IVA application for all their channels.

All applications built in Studio can access the same shared resources, including Dialogflow agents, data stores, business process automation, payment gateways, and reporting functionality. This helps ensure that IVAs maintain the context of each interaction as it progresses to the next touchpoint or escalates to a live agent.

"Recent events have demonstrated the importance of being responsive to your customers and to sudden changes that impact your business," said Callan Schebella, Inference Solutions' CEO, in a statement. "Studio 6.5 will allow any business to ramp up their self-service capabilities, increase efficiency and readiness with automation, and assist customers whenever and wherever they need it. It will also allow our telco, UC, and contact center partners to expand their offerings by supporting more of their customers' preferred channels."

Inference Studio 6.5 also includes a new Verbatim Audio feature. With this feature, raw call utterances can be used to tune speech recognition accuracy and automate more service tasks. All virtual agent tasks developed in Studio can use a variety of conversational AI engines, such as Google DialogFlow, IBM Watson and, now, Amazon Lex. Users also have access to more than 180 voices across 30 languages via the Google Text-to-Speech service, the ability to identify a voice print via ArmorVox voice biometrics, and can achieve real-time sentiment analysis using Watson Tone Analyzer.