Proving the Business Case for the Internet of Things

Oracle launches AI-Voice for Technical Use

William Payne
September 19, 2019


Oracle has launched an AI-trained voice assistant for business customers. The application will allow business workers to use domain specific language and technical terms and well as conversational language to give detailed instructions to computer systems to carry out jobs. The tool is likely to find application in busy and challenging technical environments, such as industrial and engineering, logistics and supply chain, smart building systems, and infrastructure and energy maintenance and management.

Oracle Digital Assistant is designed to apply AI with deep semantic parsing for natural language processing (NLP), natural language understanding (NLU) and custom machine learning (ML) algorithms. 

This combination allows Oracle Digital Assistant to understand a user's natural conversation, derive intent, produce compositional logical forms, and identify and learn user behaviour patterns in order to proactively take action on behalf of the user. 

A no-code tool that allows enterprises to build conversational experiences, Oracle Digital Assistant can also integrate with human agent work-flows and business processes without any coding required.

The Oracle's intelligent voice assistant analyses enterprise-specific and domain-specific vocabulary on which open and consumer-oriented domain models are not trained on. Oracle says its digital assistant can make voice and user interactions more expressive by processing complex queries and deriving intelligence from all available enterprise applications, such as ERP, CRM and HR systems to respond in the context to the request made.

Oracle Digital Assistant is pre-built with AI-trained enterprise skills across ERP, SCM, HCM and CX and can connect to multiple back-end systems simultaneously to orchestrate user interactions across various application skills. 

With these plug-and-play skills, line-of-business users only have to interface with one digital assistant that can source the right information from employee directories, expense management systems or an assortment of other enterprise applications, including Oracle Cloud Application offerings.

According to Oracle, the NLP engines that power today's traditional messaging-based channels lack the ability to handle highly expressive sentences. Voice interactions, however, enable expressive conversations which require NLP engines to manage much more complex constructs. Linguistic constructs like relative clauses, comparatives, superlatives, negation, anaphora, ordinals, cardinals, superlatives, ellipsis, quantifiers and conjunctions now need to be processed by the NLP engines that require more sophistication than the simple intent classification and slot-filling engines available today.

"Enterprises are demanding an AI-powered voice assistant that understands their specific vocabulary and enables naturally expressive interactions for its users," said Suhas Uliyar, vice president, AI and Digital Assistant, Oracle. "Most of all though, enterprises value a highly secure AI-powered voice assistant that stores their business' sensitive data in Oracle's second generation cloud infrastructure."