SUSE updated AI capabilities in Rancher Prime in time for KubeCon 2025.

The Kubernetes management system now includes a preview of its Liz context-aware AI agent and the general availability of SUSE Virtual Clusters to help optimize GPU resources.

Its SUSE Virtualization tool, meanwhile, has been refreshed to feature a beta of advanced network features such as micro-segmentation, aiming to separate network functions from physical hardware toward software-defined network configurations.

In addition, the vendor's flagship SUSE AI platform now includes the OpenTelemetry (OTel) operator for auto instrumentation, a tech preview of a model context protocol (MCP) server for agent-to-agent interactions, and a high-performance inference engine in the form of a virtual large language model (vLLM).

Speaking with SDxCentral, Thomas Di Giacomo, chief product and technology officer at SUSE, said the new release “simplifies and accelerates operations for developers and infrastructure teams.”

“For operations, the general availability of Virtual Clusters allows teams to optimize costly GPU resources. ... Our expanded Observability capabilities, including dashboarding, deliver faster incident resolution times by eliminating context switching," Di Giacomo said of the updates. “We are also introducing SUSE Virtualization advancements that provide an open, reliable path to transition away from proprietary hypervisors, increasing infrastructure agility.”

The chief exec also highlighted the introduction of Liz, an AI agent providing real-time intelligence in the operations layer of Rancher Prime to simplify the inherent complexity of Kubernetes environments.

“We’re very excited about Liz. … You might notice the name. It’s a fun nod to the SUSE family. It’s short for Lizard and a relative of our chameleon mascot, Geeko," Di Giacomo said. “Liz’s core use case is to help platform teams move beyond reactive management. Liz can proactively detect potential issues, recommend optimizations for infrastructure performance, and significantly reduce the mean-time-to-resolution (MTTR) for customer issues, freeing up engineers to focus on innovation.”

Outside of Rancher Prime, SUSE Observability has been updated with a dashboard editor that allows users to organize and display operational data more effectively. This version also introduces support for the OTel framework, allowing consistent data collection and visibility across multiple technology components, including but not limited to Kubernetes.

“OTel is truly special because it represents the open source standard for collecting, processing, and exporting all cloud native telemetry data – logs, metrics, and traces. The OTel Operator automates the deployment and management of these components,” said Di Giacomo.

The SUSE exec said OTel helped ensure vendor neutrality by decoupling data collection from analysis and allowing data to be forwarded to any compatible backend, helping prevent cloud lock-in.

“In SUSE Observability, OTel means we can provide unified visibility and insights across the customer's entire technology landscape, not just the Kubernetes environment," Di Giacomo said. "With SUSE AI, we include SUSE Observability for AI with the OTel operator for auto-instrumentation, which simplifies the deployment for applications in a cloud-native environment and is critical for bringing performance metrics and insights to AI workloads that are often difficult to monitor.”

Di Giacomo also pointed to an expanded ecosystem of certified enterprise storage providers within SUSE Virtualization, including the likes of Fujitsu, Hitachi, Hewlett Packard Enterprise (HPE), Lenovo, Dell, Portworx, and Oracle.

“This move means customers can now leverage their existing, substantial storage investments seamlessly with our SUSE Virtualization offerings," Di Giacomo said. "By supporting CSI (container storage interface) integration with these certified vendors, we are providing an open, reliable path to transition from proprietary hypervisors … simplifying the modernization journey by unifying management for both traditional VMs (virtual machines) and modern cloud-native systems on a single platform.”

In Di Giacomo’s view, the new offering reflects how organizations will operate with AI in a Kubernetes environment.

“We believe the future of enterprise IT will run on secure, containerized AI applications, and this coupling uniquely positions SUSE to accelerate our customers' AI journeys,” Di Giacomo said.