The agentic era has arrived, bringing together human insight and autonomous intelligence in entirely new ways.
The pressure is high for enterprises to become “AI-first,” but as we explored in our inaugural Agentic AI Research Report, most organizations are stuck in pilot-mode. Many are struggling to achieve meaningful outcomes, even as they invest ever more heavily on AI.
AI success requires a shift in how we think about operations. We’re moving from a world of manual changes and troubleshooting to one of autonomous systems. Observability tooling needs to adapt to this new reality. It’s not enough to merely provide insights into a system. Organizations need platforms that enable agents to act upon intelligence safely and reliably. That’s why Dynatrace has evolved into an agentic operations platform that enables autonomous, intelligent collaboration across development, operations, and business workflows.
At our flagship event Perform 2026, we’re expanding our platform to deliver the following capabilities that will carry organizations into the human + AI collaboration era.
Dynatrace Intelligence: The new foundation for reliable agentic AI
Organizations struggle with the growing complexity of cloud and AI‑driven digital ecosystems. Leaders have valid concerns about hallucinations, unreliable multi‑step agent workflows, and the inability of off-the-shelf models to process petabytes of heterogeneous observability data.
The solution isn’t to throw more AI at the problem and hope for the best. Simply adding AI agents to legacy workflows doesn’t work. Closing this gap requires an AI system that doesn’t just bolt onto workflows, but rethinks how decisions are made and actions are automated.
Dynatrace Intelligence is an agentic operations system at the core of the Dynatrace platform and serves as the reasoning and decision-making layer for all the agentic layers built upon it. It fuses deterministic AI with contextual analytics to ground agentic decisions in real‑time facts and to create more reliable agentic workflows. It also coordinates different kinds of agents while minimizing hallucinations so organizations can trust automated actions.
The new Smartscape®: A source of truth for AI
Modern IT teams often operate with only a partial understanding of what’s running in their environment, especially as cloud‑native architectures, Kubernetes, and adaptive AI agents expand and shift at runtime. You can’t troubleshoot, secure, or optimize what you can’t see. This incomplete picture slows incident response, drives misaligned investments, and increases risk.
The new Smartscape® real-time dependency graph creates a shared, real‑time source of truth that both humans and AI agents use to understand what is actually happening across the environment and to act with confidence. It provides a precise, always‑current view of every entity and dependency in your digital ecosystem, including cloud and Kubernetes objects, domain‑specific metadata, and agentless cloud‑discovered components. The new Smartscape® capability underpins Dynatrace Intelligence, providing the context needed to empower teams and agents to act with confidence, using real‑time topology to detect changes, assess impact, and respond automatically.
From reactive to proactive with agents
Far too many IT teams are forever stuck in firefighting mode. They’re so busy troubleshooting existing systems that they have no time to understand why issues keep recurring or how to prevent them. Signals flood in faster than humans can interpret them, incidents surface without context, and critical decisions must be made with only fragments of the real picture. In this chaos, teams are forced into reactive loops that drain capacity, slow innovation, and make reliable operations nearly impossible.
Dynatrace Intelligence Agents are designed to break this cycle. Whether you use our own ready-made agents or build your own, the Dynatrace Intelligence agentic ecosystem provides the capabilities to continuously detect changes, assess impact, and automatically respond to emerging conditions, helping organizations shift from firefighting to proactive operations.
This is how the Dynatrace platform goes beyond insight, advancing to autonomous action by auto-remediating, auto-preventing and auto-optimizing.
Dynatrace Assist: Your portal into the agentic universe
Modern IT teams aren’t lacking in data or metrics. The problem is that they must jump between different dashboards, tools, and consoles to understand their systems and find the information they need. This slows investigation, increases cognitive load, and leads to delays, misdiagnoses, and inconsistent decisions.
We’re launching Dynatrace Assist to eliminate this constant context switching. Assist brings the full power of Dynatrace Intelligence to your fingertips. Instead of moving between dashboards, queries, and tools, you simply start by asking a question and let Dynatrace Intelligence do the work. Assist pulls together context from the Dynatrace Grail® unified data lakehouse, maps relationships utilizing the Smartscape® real-time dependency graph, and collaborates in context with Dynatrace agents.
As AI becomes a collaborative partner rather than a separate tool, Assist provides the conversational interface that brings agentic intelligence directly into human workflows—reducing cognitive load and enabling natural, real‑time co‑reasoning.
End-to-end intelligence with next generation Dynatrace RUM
Even the smartest agentic systems need to understand how real users experience your applications. Backend signals alone can’t reveal where friction occurs, how users navigate modern single‑page and mobile experiences, or which issues matter most to the business.
The next‑generation Dynatrace Real User Monitoring closes this gap with Grail-powered analytics that unifies frontend behavior with backend intelligence. It automatically captures modern user interaction patterns and couples them with precise, real‑time system insights, enabling teams and agents to detect issues faster, prioritize what truly impacts customers, and optimize digital experiences with confidence.
Dynatrace MCP Server: The connective tissue between agentic systems and Dynatrace Intelligence
Modern AI assistants and automated workflows are being asked to make high stakes decisions without access to the complete, real-time context of the systems they’re meant to support. Fragmented data, disconnected telemetry, and incomplete understanding of how services relate prevent AI from making reliable recommendations or taking safe autonomous actions.
The Dynatrace MCP Server provides the governed bridge between AI ecosystems outside of Dynatrace and the production‑grade insights generated by Dynatrace Intelligence. Using the open Model Context Protocol, it delivers real‑time insights directly into agentic workflows. This gives autonomous agents and other AI tools the observability truth they need to reason accurately, take decisive actions, and operate safely at scale.
Building the future of autonomous operations
Perform 2026 marks a turning point for agentic AI Adoption. We’re shifting from human‑paced management of systems into a world where systems anticipate, adapt, and advance on their own. For the first time, organizations can operate with an intelligence layer that doesn’t wait to be directed but collaborates, reasons, and evolves in real time. Teams will finally be able to focus on the work that moves them forward, rather than the work that holds them back.
Looking for answers?
Start a new discussion or ask for help in our Q&A forum.
Go to forum