Enterprise software provider OpenText Corp. today announced that it will incorporate large language model-based generative artificial intelligence capabilities throughout its Information Management Cloud for content, experience, business networks, information technology operations, developer operations and cybersecurity.

The company also announced a suite of tools to aid IT administrators in creating and managing data lakes, analyzing structured and unstructured data and creating visualizations.

The LLM, called Aviator, supports natural language queries, conversational chat and personalized, automatically generated content. Organizations can use its generative capabilities within applications or via OpenText Cloud application program interface services to enable AI orchestration and information flows across multiple clouds and knowledge bases.

IT Operations Aviator is described as an AI virtual agent for OpenText’s Smax Service Management Automation platform. It combines LLMs with OpenText’s data security capabilities to facilitate self-service, provide faster issue resolution and improve efficiency in service management scenarios.

“I can log into Smax and say I’m having problems connecting my laptop to the network,” said Muhi Majzoub, OpenText’s chief product officer. “It will look at my DNS and network settings and come back with a recommendation or open a help desk ticket.”

AI for software development

DevOps Aviator applies generative AI capabilities to optimize software delivery with feature prediction and enhanced test coverage with automatic test creation and authoring. “We are not yet developing code but in the future, the goal is to see what use cases we can generate,” Majzoub said. “For now it’s test cases and risk assessment.”

Content Aviator makes generating content and customer communications more efficient by applying a chat interface and natural language queries. “Business analysts can use it to get help for a customer communication that apologizes, for example, for sending two communications in a day,” Majzoub said. “Aviator will write and present it. If it has to be rewritten to be more apologetic, it will rewrite it as needed.”

Experience Aviator adds generative AI to the company’s Customer Communications Management software to produce content that supports marketing, communications and customer service teams.

Cybersecurity Aviator uses AI to detect a variety of cybersecurity threats. Network Aviator applies generative AI to supply chain information flow to give businesses a conversational interface for accessing information related to market changes.

Enhanced IT support

OpenText Aviator Platform supports IT organizations with enhanced search, endpoint management and developer tools. Aviator Search spans all data types across multiple repositories for use in building custom applications and portals. “We have 68 million documents in our internal document management system and it can make recommendations or identify documents related to a specific customer contract,” Majzoub said. “It can be used against OpenText document management platform in any industry.”

Aviator IOT is a collection of tools to better connect and protect large numbers of internet of things endpoints with real-time visibility into assets’ location, condition, utilization, performance and health. Majzoub used the example of a pharmaceutical company that needs to track cold shipments using sensors and global positioning system coordinates.

“The Aviator can track that truck number seven going from Pennsylvania to Florida has a high temperature that will ruin the drugs unless the problem is corrected,” he said. “It can notify administrators or truck drivers to resolve the problem immediately.”

Aviator Thrust & Thrust Studio is a set of cloud API services and developer tools the company built over the last three years to support secure information flows, custom AI development and AI-embedded applications. Aviator Lab accelerates AI development through rapid prototyping of AI reference architectures in a secure sandbox environment.

Majzoub said OpenText will work with customers to custom-train LLM models without the risk of data leakage. “Customers will bring data into a private environment that only they have access to and the model is private to them,” he said. “They can experiment and decide whether to subscribe in production or turn any feature off.”

Most new features will be available in OpenText version 23.4, which will ship by the end of this month. Some will be available in version 24.1, which will be released in the first quarter of next year.

Photo: OpenText