Juniper Networks Inc. today unveiled significant upgrades to its Mist networking platform, introducing new capabilities that extend insight and automation across wired, wireless and wide-area network environments.

The updates to the artificial intelligence-native platform include an expanded version of Marvis Minis for digital experience twinning across global cloud infrastructure, a new Marvis Actions dashboard for autonomous network control, and an upgraded Marvis Client that extends AI-native operations to end-user devices. The updates are aimed at moving beyond traditional observability tools by leveraging machine learning and telemetry to automate both insight and remediation.

Marvis Minis

Up first is an update to Marvis Minis, Juniper’s AI-driven digital experience twins in Juniper’s Mist platform. They proactively simulate and monitor end-to-end user experiences across the network to detect and resolve issues before they impact performance.

The updated Marvis Minis expands Juniper’s digital twin concept across the entire client-to-cloud journey, using AI to detect and analyze performance issues across applications, internet service providers, sites and regions. The update introduces new service-level expectations to benchmark application responsiveness and user experience in real time. The AI-native approach replaces the need for traditional agents or sensors and delivers a seamless and scalable method for maintaining service quality.

Juniper says this capability enables proactive remediation of problems before users are affected. Marvis Minis can isolate performance degradation to specific areas, be it a cloud application, a local switch or an ISP bottleneck, all by running continuous synthetic tests.

Marvis Actions

As part of the update, Juniper has introduced the Marvis Actions dashboard, a new dashboard that gives network teams visibility and control over automated decisions made by the Marvis AI Assistant. The dashboard allows operators to choose how and when AI should take action on issues such as virtual local area network misconfigurations, radio interference or outdated firmware. Each action is documented in a detailed activity log along the way, allowing for full auditability and human oversight.

The feature aligns with Juniper’s vision of a “self-driving network,” where common maintenance tasks and performance optimizations are carried out automatically. The result is reduced operational overhead and increased responsiveness, particularly in highly distributed or complex environments, with customers getting greater assurance and flexibility without compromising visibility or control.

Marvis Client

The third and last announcement, the Marvis Client extension, brings telemetry collection directly to user devices running Android, Windows or macOS, capturing data such as operating system details, hardware specs, signal quality and connectivity issues. The data is then processed by the Mist cloud and correlated with insights from Juniper access points, switches and firewalls to provide a unified view of the end-user experience.

The client-based insight does not require additional sensors or third-party software, making it a low-cost, high-value enhancement for organizations seeking deeper observability. The outcome helps reduce mean time to resolution and improves information technology team productivity, especially in hybrid and remote work environments.

“The Mist AI-native networking platform was purpose-built to converge AI and networking for exceptional operator and end user experiences,” said Sudheer Matta, senior vice president of products, campus and branch at Juniper. “These enhancements shift the paradigm from traditional observability to an AI-native model for truly understanding user experience that’s actionable at scale.”

Neil McRae, chief network strategist at Juniper Networks Inc., spoke with theCUBE, SiliconANGLE Media’s livestream studio in March to discuss how Juniper is aiming to simplify complexity with AI-native networking:

Photo: Juniper Networks