Agents for the network admin: New AI features from HPE Juniper

Following the acquisition by Hewlett Packard Enterprise (HPE), Juniper has expanded its Mist portfolio with new features that drive the use of agent-based, AI-powered network automation (AIOps). The aim is to make network operations increasingly autonomous and to move from reactive to proactive operating models.

One focus is on expanding the AI assistant Marvis, which now provides improved conversation functions for real-time error analysis. With the help of agentic workflows, data from different network areas is brought together to enable precise problem solutions. In parallel, the Marvis Actions dashboard now supports a wider range of automated corrective actions, such as port misconfigurations, capacity bottlenecks or non-compliant hardware. With the Large Experience Model (LEM), HPE is also introducing an analysis model that evaluates telemetry data from applications such as Zoom or MS Teams in order to detect performance problems at an early stage. In conjunction with the so-called Marvis Minis, which simulate user experiences, future quality losses can be predicted and rectified before they occur.

The AIOps functions have also been expanded for data centers: Marvis AI is linked to Apstra's context database and thus provides deeper insights into data center networks. In addition, the Marvis Minis enable continuous validation of services and safeguarding of applications in data center operations. These innovations are embedded in the HPE GreenLake Intelligence platform, which provides AI agents in a multi-layered IT architecture to support real-time decision making, optimization and problem solving across network, storage and compute resources. The new Marvi data center capabilities complement OpsRamp, an AIOps-based platform for IT operations management of hybrid and multi-cloud environments with enhanced visibility and automated workflows.

Read also

HPE and Juniper: The path to the new number one in the network market?

(dmk)