Dynatrace Drives Business Observability for Agentic AI

Dynatrace is tackling the complexities of autonomous agentic AI with a new focus on business observability, ensuring these intelligent systems achieve their goals within defined guardrails.

The technological landscape is undergoing a seismic shift, one where the very fabric of enterprise operations is being rewoven by the invisible threads of artificial intelligence.

What began as a cautious exploration of AI models has rapidly evolved into a full-blown embrace, pushing the boundaries of what businesses thought possible and, in turn, demanding a revolutionary approach to understanding and managing these intelligent systems.

At the forefront of this evolution, companies like Dynatrace Inc., traditionally stalwarts in observability and security for conventional workloads, find themselves uniquely positioned to chart this new frontier, now extending their gaze to the intricate world of AI.

Alois Reitbauer, Chief Technology Strategist of Dynatrace, offers a compelling perspective from this vantage point.

He observes a profound transformation in how applications are constructed, a shift he describes as moving beyond the initial reliance on single, dominant models like OpenAI.

“In the past it was basically OpenAI, you used OpenAI and then it started to switch to other models.

Now, we see people experimenting way more, like A-B testing models and the practice of… AI native engineering,” Reitbauer explained during a recent interview with theCUBE at KubeCon + CloudNativeCon NA.

This isn’t just about picking a different flavor of AI; it signifies a deeper commitment to integrating AI at the core of business processes, fostering an environment of continuous experimentation and refinement.

Enterprises are no longer merely adopting AI; they are becoming “AI native,” building their systems from the ground up with intelligence embedded at every layer.

This rapid, almost frenetic, adoption of diverse AI models across the enterprise has thrown a spotlight on a burgeoning challenge: how do you maintain visibility and control over systems that are increasingly autonomous?

While individual AI models are making strides in transparency – often capable of sharing a “train of thought” that illuminates their decision-making process, a welcome development for debugging – the rise of agentic AI introduces a whole new stratum of complexity.

Agentic AI refers to autonomous software agents designed to achieve specific goals, often by orchestrating multiple AI models and tools.

They don’t just execute commands; they perceive, reason, plan, and act independently to achieve their delegated tasks.

Reitbauer highlights this paradox: “One of the very positive trends was that AI models now actually share a train of thought that helps you to understand how the model arrived at where they arrived.

That helps us a lot on the debugging side.

But debugging AI in agentic applications is kind of different.

The more we move into more dynamic systems, like going more into this agentic world, the more the individual transactions will be different.”

The implication is profound.

While we might understand the internal workings of a single AI component, the emergent behavior of a system of interconnected, self-directing agents becomes incredibly difficult to predict and monitor using traditional methods.

Each interaction, each “transaction,” becomes a unique journey through a labyrinth of decisions and actions, rendering static monitoring tools largely ineffectual.

This is where Dynatrace sees its mission evolving.

The company recently announced a next-generation cloud operations solution for Microsoft Azure, specifically designed to support the platform’s agentic AI initiatives.

The focus, Reitbauer emphasizes, is on bringing a new form of observability to these agentic operations.

The traditional paradigm of “looking AI over the shoulder,” meticulously tracking every computational step, simply doesn’t scale or make sense when you’re dealing with an agent delegated to achieve a complex task.

Instead, the imperative shifts to “business observability.”
“Guardrails are a key and guardrails started to emerge very early on,” Reitbauer noted.

“Really thinking to the next step about agentic, we have to track against goals and I think that’s where business observability comes in.

You’re delegating a task, you’re not looking AI over the shoulder.”

This represents a fundamental reorientation.

Rather than scrutinizing the minutiae of AI execution, enterprises must now focus on whether the autonomous agent is successfully achieving its assigned business objective within predefined boundaries.

Are the sales agents meeting their quotas?

Is the customer service bot resolving issues efficiently and ethically?

Is the supply chain optimizer reducing costs without compromising quality?

The “how” becomes less important than the “what” and the “whether.”

This paradigm shift necessitates a robust framework of guardrails – ethical, operational, and performance-based – to ensure that autonomous agents operate within acceptable parameters.

It demands a sophisticated observability layer that can translate the complex, dynamic actions of AI agents into actionable insights tied directly to business outcomes.

The challenge for enterprises isn’t just deploying AI; it’s about deploying responsible and accountable AI.
The journey into agentic AI is still in its nascent stages, yet its implications are already reshaping enterprise strategy.

It promises unprecedented levels of automation and efficiency, but also introduces a new frontier of operational complexity and ethical considerations.
For companies like Dynatrace, the task is clear: to provide the navigational tools for this new era, ensuring that as AI agents gain more autonomy, human operators gain more clarity and control, ultimately fostering trust in the intelligent systems that are increasingly running our world.

The future of enterprise AI isn’t just about building smarter machines; it’s about building smarter ways to observe, understand, and govern them.