2 min

Tags in this article

, , , ,

HPE is integrating several self-developed LLMs into its HPE Aruba Networking Central platform to make network environments running on it easier to manage. These LLMs not only provide the company’s own technical data but also discover proprietary and personal data.

In particular, the tech giants added self-developed and trained LLMs to bring new GenAI capabilities to the search function within the cloud-based network management platform. These capabilities complement the already existing AIOps functionality for troubleshooting assistance and other technical information, HPE indicates.

Functionality

The new LLMs focus primarily on making using the cloud-based network management environment easier. The search function, located at the top of the interface, can now better understand queries that contain network technical terms.

Een interface van hpe greenlake die een netwerkbeheerdashboard toont met een globaal overzicht van verschillende netwerksites en hun status.

A new autocomplete feature presents users with multiple queries based on the text they enter. Results also come back faster than those of other chatbots, HPE claims.

In addition, LLMs provide functionality that summarizes documentation about HPE Aruba networking solutions and applications. Questions about tasks are answered in plain language based on the relevant technical guides. Includes links to the documents from where the information was extracted.

Proprietary and personal data

One of the new HPE LLMs is also designed to discover proprietary and personal information. As a result, the search function better understands queries about company-specific issues, such as office location. Second, it can discover specific LLM queries with personal data and prevent these data from being included in the LLM’s training sets.

The new LLMs are trained with data models 10 times larger than those of competitors, according to HPE. Among other things, more than 3 million questions that end users have asked Aruba Central have been added to the models. In addition, the models were trained with tens of thousands of technical documents published by HPE for its Aruba network portfolio.

The LLMs will gradually become available in Aruba Central this month. The entire rollout of the new GenAI functionality should be complete by April.

Also read: HPE Aruba is working on “Aruba Central Next Gen”: what is that?