Remediate without hallucinations using Time-Traveling Topology with SUSE Observability’s AI Agent 35+ MCPs

At KubeCon EU 2026, SUSE is announcing AI assisted incident investigation in SUSE Observability. The new capability helps platform teams reach root cause faster by combining Agentic AI with Time Traveling Topology, a model that captures the full state of the system at every moment in time.

Instead of guessing across dashboards, the platform compares the system before and during an incident to identify exactly what changed across services, infrastructure, and dependencies. Because the AI is anchored to Time Traveling Topology, it operates on real system state rather than probability. This dramatically reduces AI hallucinations and gives engineers answers they can trust. It also reduces the time teams spend trying to understand what happened, a challenge many teams now call Mean Time to Context.

As environments grow across clusters, clouds, and edge locations, that context becomes harder to reconstruct manually. Dependency chains stretch across hundreds of services and workloads appear and disappear constantly. During an outage, rebuilding that picture by hand is nearly impossible.

Your AI Observability likely falls short

Many observability tools now promise AI assistance. Most attach an LLM to logs and dashboards and call it a copilot.

But these systems lack the full context of the infrastructure. They analyze telemetry in isolation and make statistical guesses about what might be related. They do not understand the real dependency graph of the system or what the environment looked like before the incident began.

The result is confident sounding recommendations that are often wrong. During an outage that wastes the one thing teams do not have: time.

A different model built on Time‑Traveling Topology

SUSE Observability approaches the problem differently. The platform continuously captures a versioned snapshot of the entire system topology. Every service, dependency, and health state is preserved as a point in time model.

Engineers or the integrated Agentic AI troubleshooting can rewind the environment to any moment before or during an incident and see exactly how the system looked at that time. Even components that no longer exist, such as terminated pods or rescheduled workloads, remain visible in the historical topology.

This becomes the foundation for AI assisted investigation.

Instead of guessing, the AI compares the healthy system state to the state at the moment the anomaly began. It analyzes the full dependency graph and all other signals via the integrated 35+ MCP tools to highlight the precise components that changed.

The investigation becomes deterministic rather than speculative. Because the AI analyzes versioned topology snapshots instead of isolated telemetry, it avoids the hallucination problem common in many AI observability tools.

AI Investigation with the Evidence

When an incident occurs the AI agent automatically captures the topology at the start of the anomaly and compares it with the previous healthy state. It evaluates upstream and downstream dependencies and correlates events, metrics, logs, and traces within that full system context.

The output is not a guess. It is a clear chain of evidence that shows what changed and where the failure propagated.

For engineers this dramatically reduces the time required to understand an incident. Instead of jumping across tools and dashboards, teams start the investigation with a complete picture of the system.

The SUSE Observability AI Agent Joins the Rancher AI Crew

The SUSE Observability AI agent is also part of the new AI agent Crew in SUSE Rancher Prime also announced at KubeCon EU 2026. The AI Crew is an agent ecosystem embedded in SUSE Rancher Prime with specialized agents for observability, security, Linux operations, provisioning, and fleet management.

The observability agent does not work alone. It can pull context from the other agents in your stack. A recent security policy change, a fleet rollout, or a node level issue can all become part of the investigation.

The result is cross domain context that traditional observability tools cannot provide.

See it in Action

[YouTube Video Embed]

Extend the investigation with MCP

Through Model Context Protocol support in Rancher Prime, organizations can also connect their own operational systems to the AI Crew. Internal CMDBs, ticketing systems, or security scanners can contribute context to the investigation without custom integrations.

This allows the AI to reason across the entire operational environment instead of a single tool.

Why this matters for platform teams

When teams reach the root cause faster, incidents shrink. Engineers spend less time switching between tools and more time fixing the problem.

Topology aware AI also reduces cognitive load. The system performs the cross correlation work that previously required deep tribal knowledge of the platform.

For leaders the outcome is simple. Faster investigations, fewer escalations, and lower operational risk across distributed infrastructure.

Available now

AI assisted investigation is generally available in SUSE Observability starting with SUSE Rancher Prime v2.14. It is available on premises with Rancher Prime and will become available through SUSE Observability Hosted, the fully managed hosted offering and we have plans of extending availability soon.

If you are attending KubeCon EU in Amsterdam, visit the SUSE booth to see live demonstrations of Time Traveling Topology and AI investigation across a real multi-cluster environment.

Stop guessing during incidents

Every minute spent searching for root cause increases customer impact and engineer fatigue. Observability should shorten that investigation, not extend it.

SUSE Observability combines topology aware telemetry with AI agents that understand the full system context. The result is faster investigations and higher confidence in the answer.

Stop guessing. Start knowing. 

See these capabilities in action at KubeCon EU 2026 in Amsterdam, or connect with a SUSE expert to explore what’s possible for your organization

Come see us at KubeCon EU in Amsterdam.

Visit the SUSE booth, join our sessions, and experience firsthand what an AI-native cloud native platform can do for your organization.

For the latest updates, visit suse.com/kubecon and follow us on social media throughout the week.