Research Finder
Find by Keyword
Can AI Agents Truly Replace Human IT Operators?
Dynatrace shifts from passive monitoring to autonomous operations using deterministic AI agents, no-code builders, and broad multi-cloud integrations.
7/29/2026
Key Highlights
- Dynatrace is shifting its focus from passive monitoring to active issue resolution.
- The introduction of the Autonomous SRE Agent aims to automate incident triage.
- A new no-code Agent Builder empowers teams to create custom automation workflows.
- The Cloud SRE Agent centralizes multi-cloud remediation into a single auditable record.
- We see this as a move to ground autonomous actions in deterministic facts rather than guesses.
The News
Dynatrace has announced significant updates to its Intelligence platform that combine autonomous agents with deterministic system mapping to automatically resolve IT incidents. The new capabilities include an Autonomous SRE Agent, a Cloud SRE Agent, and a no-code Agent Builder designed to extend automation across varied workflows. These updates aim to provide enterprises with reliable execution and human oversight for complex multi-cloud environments.
Analyst Take
For years, the technology industry has promised that AI would solve the sheer complexity of modern infrastructure. We see teams drowning in alerts, logs, and dashboards on a daily basis. The burden on operators is immense. Yet, the initial wave of AI tools often fell short of expectations. They relied heavily on probabilistic models that excel at generating text but struggle with the strict precision required to manage enterprise systems. When a core banking application goes down, guessing the root cause is not an option. Accuracy is absolutely essential. Enterprises need facts, not assumptions.
This latest announcement from Dynatrace aims to deliver a different approach to this persistent problem. By focusing on deterministic context, the company is attempting to bridge the gap between generating insights and taking safe, governed actions. The premise is straightforward but technically demanding. Grounding an autonomous agent in a real-time, factual map of the environment makes safe execution more achievable, but context alone does not create trust. Enterprises also need policy-aware execution, least-privilege access, defined escalation thresholds, and the ability to stop or reverse an action when the system behaves unexpectedly. This shifts the role of the observability platform from a passive dashboard that human operators must interpret into an active participant in the resolution process. It acts on answers.
We view this transition as highly necessary. Modern cloud environments change by the minute. According to HyperFRAME Lens data, 79% of public and hybrid enterprises now use multiple cloud providers, with 27% running four or more, creating significant architectural sprawl and operational complexity. Human operators simply cannot keep pace with the sheer volume of telemetry data generated across microservices, containers, and serverless functions. Automation is truly essential. However, trust remains the primary barrier to adoption. Trust is hard won. Operations teams are understandably hesitant to hand over the keys to an AI system without robust safeguards and auditable trails. Dynatrace is architected to address this hesitation directly by ensuring that every automated action is logged, transparent, and subject to human oversight when required.
The critical issue is not whether an agent produces a factually grounded recommendation. It is whether the platform can govern what happens next. An agent that correctly identifies a problem can still choose an inappropriate remediation, affect a dependent system, or operate beyond its intended authority. Observability, policy enforcement, identity, approval controls, and rollback therefore need to function as one operational system. This is where observability platforms are beginning to evolve into runtime control planes for agentic operations.
So can AI Agents truly replace human IT operators? Our answer is no, not entirely, and certainly not yet. AI agents can replace significant portions of repetitive investigation and remediation, but they do not eliminate the need for human operators. They change where humans add value: from interpreting every alert and executing every fix to setting policies, defining escalation thresholds, and supervising higher-risk actions.
What Was Announced
Dynatrace announced several major enhancements to the Dynatrace Intelligence platform that are designed to facilitate autonomous operations. At the core of this release is the Autonomous SRE Agent. This agent is architected to trigger autonomously the moment a new problem is detected within the environment. It is designed to evaluate whether the new issue is related to an ongoing investigation or if it represents a novel incident. If it finds a connection, the agent enriches the existing investigation with additional technical context and updates the problem record. This capability is expected to be available in August and aims to deliver a significant reduction in duplicate alerts and manual triage efforts.
Alongside the Autonomous SRE Agent, the company introduced the Cloud SRE Agent, which is available to SaaS customers immediately. This tool is designed to coordinate remediation activities across major hyperscalers, integrating directly with existing agents across AWS, Microsoft Azure, and Google Cloud environments. It aims to deliver a single, centralized, and auditable record of all autonomous operations. By consolidating these findings, the Cloud SRE Agent is architected to provide operations teams with a clear view of what actions were taken, why they were taken, and what the resulting impact was on the infrastructure.
Furthermore, Dynatrace introduced the Agent Builder, a no-code environment also slated for an August release. This feature is designed to allow customers to construct and deploy custom AI agents tailored to their specific operational workflows. We see this as a highly practical addition. No two enterprise environments are exactly alike, and out-of-the-box automation rarely covers every edge case. By enabling teams to build bespoke agents without writing code, Dynatrace aims to democratize access to autonomous operations, allowing domain experts to codify their knowledge into repeatable, automated actions.
The release also includes Enhanced Dynatrace Assist, which brings natural language investigation capabilities and agent-ready workflows to a broader set of users. This is designed to lower the barrier to entry for querying complex system data. Finally, the company announced an expanded integration ecosystem. New integrations with enterprise platforms like ServiceNow, Atlassian, and PagerDuty are designed to ensure that Dynatrace Intelligence can orchestrate remediation across the systems that IT teams already rely on every day.
By combining these new features, the platform is architected to move organizations beyond mere visibility. The focus is squarely on automated execution. We believe that grounding these agents in a deterministic understanding of the underlying infrastructure is a sound strategy. It mitigates the risks associated with probabilistic AI and provides the auditability that enterprise governance teams demand. The success of these tools will ultimately depend on how easily customers can configure them to match their unique operational maturity levels.
We see a broader implication for the industry here. For a long time, the burden of action rested entirely on human shoulders. The software simply pointed out that a problem existed. Now, the software is designed to take the next logical step. It analyzes the root cause, formulates a remediation plan, and executes the fix. This transition requires a fundamental shift in how IT operations teams organize their daily work. They will need to move from being firefighters to becoming supervisors of intelligent systems. They will monitor the agents that monitor the systems. It elevates human operators.
We find the integration strategy particularly noteworthy. A common pitfall for observability vendors is attempting to force customers into a single, closed ecosystem. By expanding integrations with major incident response and service management platforms, Dynatrace acknowledges the reality of modern enterprise IT. Tools must play well together. The Cloud SRE Agent is architected to act as a coordinator across these various domains rather than a replacement for them. This pragmatic approach aims to deliver faster time to value for organizations that have already invested heavily in their existing operational toolchains. We will be closely watching how these integrations perform in complex, high-volume production environments.
Looking Ahead
Based on what we are observing, the transition from passive system monitoring to autonomous execution represents a fundamental realignment of the enterprise IT operational framework. The key trend that we are going to be looking out for is how effectively organizations can synthesize these deterministic AI agents into their existing incident management protocols without compromising stringent governance standards. Our perspective is that the market is rapidly moving past basic telemetry aggregation; vendors that cannot offer verifiable, auditable automation will soon find themselves relegated to legacy status.
We do not believe the near-term end state is fully autonomous IT operations. Enterprises are more likely to adopt graduated autonomy: allowing agents to resolve low-risk, repeatable incidents while requiring human approval for actions with a larger operational, security, or compliance impact. The winning platforms will not be those that remove humans fastest, but those that help organizations expand agent authority safely as trust is earned.
The announcement positions Dynatrace aggressively against key competitors like Datadog, Splunk, and New Relic. While Datadog continues to excel in frictionless developer adoption and Splunk maintains a stronghold in security information and event management, Dynatrace is attempting to carve out a distinct moat built on causal AI and deterministic automation. Going forward, we are going to be closely monitoring how the company performs on user adoption metrics for the no-code Agent Builder. The ability for non-engineers to construct bespoke operational agents could significantly alter the total cost of ownership for observability platforms.
However, lowering the barrier to agent creation also lowers the barrier to agent sprawl. The enterprise value of Agent Builder will depend on whether Dynatrace can provide centralized discovery, testing, permissions, versioning, and lifecycle governance for the agents users create. No-code makes automation more accessible; it does not make that automation inherently safe.
HyperFRAME will be tracking how the company manages the rollout of the Autonomous SRE Agent in future quarters. The theoretical benefits of automated remediation are substantial, yet practical execution within highly regulated environments remains the ultimate test of efficacy. This operational friction is reflected in HyperFRAME Lens data, which shows that 84% of organizations report AI deployments consuming more budget and operational resources than planned, while only 23% of enterprise AI projects successfully reach production and meet original ROI objectives. We see this evolution as a necessary maturation of the sector, demanding rigorous validation from enterprise architects.
Stephanie Walter | Practice Leader - AI Stack
Stephanie Walter is a results-driven technology executive and analyst in residence with over 20 years leading innovation in Cloud, SaaS, Middleware, Data, and AI. She has guided product life cycles from concept to go-to-market in both senior roles at IBM and fractional executive capacities, blending engineering expertise with business strategy and market insights. From software engineering and architecture to executive product management, Stephanie has driven large-scale transformations, developed technical talent, and solved complex challenges across startup, growth-stage, and enterprise environments.
Steven Dickens | CEO HyperFRAME Research
Regarded as a luminary at the intersection of technology and business transformation, Steven Dickens is the CEO and Principal Analyst at HyperFRAME Research.
Ranked consistently among the Top 10 Analysts by AR Insights and a contributor to Forbes, Steven's expert perspectives are sought after by tier one media outlets such as The Wall Street Journal and CNBC, and he is a regular on TV networks including the Schwab Network and Bloomberg.