Author: Chris Farrell

Turn Observability into Action: IBM Instana and Red Hat Ansible Automation Platform for AIOps

At Red Hat Summit 2026, I had the opportunity to speak on an AIOps panel in the Red Hat Ansible Automation Platform product spotlight where we covered the full AIOps journey: from first signal to automated resolution to post-incident analysis.

The framing that opened the session captured what we hear from customers every day: observability platforms are surfacing more signals, more context, and more intelligence than ever. AI is accelerating that trend. The question is no longer “what's happening?” It is “what happens next with all of that intelligence?”

Many organizations already have all the individual pieces: playbooks they trust, workflows with change management, and Observability platforms generating the right signals. And connecting them all is the execution layer. That is where Red Hat Ansible Automation Platform fits. With multiple ways to interact with the platform including APIs, event-driven automation, web UI, and Model Context Protocol (MCP), teams can select whichever is appropriate for their scenario and maintain the same governed execution environment.

Breaking down complexity with Instana automated observability

You can’t overcome complexity issues across modern environments by simply better monitoring. Context is a bigger problem to overcome.

Instana’s single-agent architecture auto-discovers over 300 different platform, runtime and infrastructure technologies with zero configuration. You do not tell it what is running, or how things are connected.. We find every running entity, deploy and configure monitoring, map dependencies, and start tracing every request immediately. That gives SREs, platform engineers and IT operations teams a live picture of how services, infrastructure, and deployments connect, without weeks of manual instrumentation.

The latest innovation that helps is Agentic causal AI. When something degrades, Instana does not stop at "something is wrong." We trace the probable root cause across the full stack. Adaptive thresholds learn your environment's normal patterns, including seasonality, so you alert on real anomalies, not static thresholds that generate false positives.

That context is the difference between an alert and an action plan. Instana tells Ansible Automation Platform not just that there is a problem, but what the problem is, what changed, and where to start. The right playbook can fire without an engineer spending twenty minutes (or longer) on triage.

Building confidence and efficiency Into AI-driven operations

There is a gap every IT leader feels: wanting AI to act and trusting it to act. The other panelists and I agreed on the concept that confidence is earned, not declared.

I recommend a disciplined progression: start with people in the loop, then on the loop, then eventually supervising at the loop. When organizations first connect Instana to Ansible Automation Platform, start with a known failure pattern such as a service that degrades the same way every quarter. Put an approval gate in the middle. The human still signs off.

After thirty to sixty days of accurate detection and correct remediation, teams naturally ask: "Why are we still approving this manually?" That progression from supervised to autonomous is how trust gets built in production.

Instana's adaptive thresholds and causal AI provide confidence in the signal. Ansible Automation Platform's governance, which includes role-based access control (RBAC), approval workflows, and audit trails, provides confidence in the action. Every automated response runs through the same execution layer with consistent controls, logging, and accountability. That consistency is what lets organizations expand automation safely.

Where to start

Even the best AI is susceptible to the problem of “garbage in, garbage out.” Before you automate, invest in data richness: high granularity, high cardinality, and a dependency model that shows how everything connects.

Instana delivers that automatically. Once you see your environment clearly, pick one well-understood failure pattern, and wire it to an Ansible Automation Platform playbook your team already runs manually. Do not start by building new automation from scratch. Start with an Ansible Automation Platform playbook you already trust.

That first closed loop (even with an approval gate in the middle) changes the entire conversation about what is possible. It gives you the audit evidence to justify expanding, and it proves the model before you scale it.

A real world AIOps, observability and action use case

A common use case we see is service latency spike recovery, and it is among the most costly when handled manually. In most organizations, latency spikes sit in alert queues waiting for an on-call engineer to notice, diagnose, and respond. Every minute in that queue extends the customer impact window.

Here is how the joint workflow changes that:

1. Detect. Instana's adaptive thresholds catch the latency anomaly in real time, using learned baselines rather than static thresholds to separate genuine degradation from normal variation.

2. Diagnose. Instana’s Causal AI traces the probable root cause across services, infrastructure, and recent deployments, giving teams the context needed to select the right response.

3. Act. The moment the threshold is breached, Ansible Automation Platform executes a governed remediation automatically. A pre-tested playbook runs the proven recovery steps your team would perform manually: no ticket queue, no manual triage, no waiting to get paged.

4. Validate and document. When remediation completes, Ansible Automation Platform posts validation and annotation back to the Instana incident timeline and includes what ran, when, and whether the service recovered.

Figure 1: Architectural diagram: Service latency and spike recovery

The result: service degradation resolves in minutes instead of hours, without human intervention. Every action carries a full audit trail, making the response consistent across shifts and compliant with governance requirements.

Key takeaways

●      The AIOps opportunity is not about better alerts. Organizations must turn observability intelligence into governed, auditable, and repeatable action.

●      Instana brings zero-config discovery, causal AI, and adaptive thresholds so teams act on real problems, not noise.

●      Red Hat Ansible Automation Platform is the execution layer that makes Instana's context actionable with RBAC, approvals, and audit trails built in.

●      Trust in automation is earned over time: start supervised, prove results, then expand.

●      The fastest win is one known failure pattern, one trusted playbook, one closed loop.

Next steps

If this resonated with you, explore the following assets to learn more:

●      Watch the panel recording.  Hear the full discussion on connecting observability, ITSM, and governed automation across the AIOps stack.

●      Try this interactive walkthrough.

●      Read the solution guide for step-by-step guidance on joint use cases.

Red Hat Summit 2026: Ansible Automation Platform product spotlight AIOps panel