Networks keep getting more distributed and complex. As companies expand across cloud, edge, and hybrid environments, IT teams are tasked with keeping everything resilient, efficient, and user-friendly. HPE is tackling this with new agentic AI-native capabilities in its Juniper Networking portfolio, aimed at making network operations more autonomous, proactive, and data-driven.
Agentic AI for Smarter Network Operations
These updates build on the
Mist AI platform, which already drives HPE Juniper Networking’s self-driving operations. Mist collects telemetry across wired, wireless, WAN, and data center networks, automates workflows, and connects with tools like Zoom, Teams, and ServiceNow to speed up root-cause analysis. The new features push this further with agentic AI - specialized agents within an AIOps framework that can detect, predict, and fix problems before they disrupt users.
Jeff Aaron, Vice President of Networking Product & Solution Marketing at HPE described self-driving operations as a staged journey rather than an immediate transition.
"Agentic AI and autonomous networking improve productivity, streamline operations, and improve efficiency but never at the expense of human oversight. The expanded self-driving actions dashboard uses embedded human-in-the-loop controls so operators can enable autonomy at their own pace. All autonomous actions are summarized in efficacy reports, ensuring IT teams maintain full visibility and the ability to approve, monitor, or revert changes across even the most complex multi-vendor environments," Aaron told ChannelE2E.
Key advancements center on expanding the Mist AI platform’s ability to anticipate and resolve issues across the network. The Marvis AI assistant now offers enhanced conversational troubleshooting, providing real-time insights through agentic workflows that span wired, wireless, WAN, and application domains. The Marvis Actions dashboard extends its reach as well, autonomously remediating more problems - such as misconfigured ports or capacity shortfalls - while preserving IT oversight. HPE has also introduced the Large Experience Model (LEM), which ingests billions of collaboration app data points and, when paired with Marvis Minis as digital twins, can forecast application behavior and resolve potential performance issues before they impact users. For data centers, new integration with Apstra’s contextual graph database allows the Marvis AI Assistant to deliver intelligent insights and continuous service validation, laying the foundation for autonomous provisioning and end-to-end assurance.
Aaron highlighted the difference between HPE’s approach and traditional telemetry-driven models: “Marvis Minis act as digital twins that synthetically test the network and correlate historical app data with real-time conditions to flag potential issues before users are impacted. By ingesting this data, the Large Experience Model becomes a generalized LLM for IT operations. Its effectiveness shows up in fewer bad user minutes, faster root-cause identification, and improved uptime across collaboration platforms.”
Extending Across the Full IT Stack
These innovations feed into HPE’s broader GreenLake Intelligence strategy, which applies agentic AIOps across networking, compute, storage, and applications. By embedding AI agents at multiple layers of the IT architecture, GreenLake Intelligence enables real-time optimization and proactive problem-solving across the entire infrastructure.
According to Aaron, this convergence creates new opportunities for partners: “User experience insights and performance rely on cross-domain, client-to-cloud telemetry. GreenLake Intelligence unifies networking, compute, and storage with a secure AI-native platform built on a single AI engine and centralized operations. It delivers client-to-cloud agents with domain expertise, enabling partners to offer outcome-based managed services across a broad spectrum of solutions.”
The new Marvis capabilities for data centers also complement HPE OpsRamp, an AIOps-powered IT operations management platform. Together, they create a more unified, full-stack approach to observability, automation, and agentic workflows for hybrid and multi-cloud environments.
Preparing for the Next Era of Networking
HPE has been building toward AI-native, self-driving operations for more than a decade. With the latest Mist updates, the company is advancing from reactive network management to proactive, predictive intervention. For enterprises, service providers, and telcos, this means faster resolution, improved efficiency, and greater assurance of user experiences from client to cloud.
As Aaron noted, partners are central to this shift. “Partners can now monetize self-driving operations by offering AI-powered assurance, predictive troubleshooting, and automation-as-a-service. We’re making it easier to sell, deploy, and support these capabilities with simplified contract addendums, cross-portfolio training, and access to the full suite of Mist and Aruba tools.”
