Dynatrace Intelligence Fuses Agentic AI with Determinism for Autonomous Operations

Dynatrace Intelligence Addresses Critical Agentic AI Failure Rates

At Dynatrace Perform 2026 in Las Vegas, the company unveiled Dynatrace Intelligence, marking a significant evolution in how observability platforms handle autonomous software operations1

1

. CTO Bernd Greifeneder exposed a fundamental problem plaguing agentic AI implementations: when an LLM with 95% accuracy executes 10 sequential calls to solve complex incidents, error rates accumulate and success drops to roughly 60%. This compounding failure rate makes most agentic AI systems unreliable for production environments where IT operational observability demands precision.

Greifeneder demonstrated why determinism must precede any LLM interaction: "The better structured the information is, the more context you have, the more causal information you provide to an AI, the better the outcome is"1

1

. The challenge intensifies when dealing with petabytes of logs that burst context windows and increase hallucinations risk. Modern observability scenarios illustrate this complexity—a degraded checkout flow might trace back to database contention from unrelated analytics queries, requiring AI agents to traverse causal chains without losing accuracy at each inference step.

Deterministic and Agentic AI Architecture Delivers 12x Performance Gains

Dynatrace Intelligence addresses these limitations by deploying three deterministic AI agents that establish factual grounding before generative AI enters the workflow1

1

. The root cause analysis agent examines millions of causal dependencies using Dynatrace's Smartscape topology graph. An analytics engine transforms exabytes of data stored in the company's Grail data lakehouse into context-rich information optimized for AI consumption2

2

. A forecasting agent scales predictive analytics across millions of metrics simultaneously.

Benchmark results show 12 times higher success rates, three times faster incident resolution, and half the token costs compared to LLM-only approaches1

1

. Greifeneder noted that "the bigger, the more complex your environment is, the better those numbers get"1

1

. The system delivers trustworthy, explainable insights that AI agents can safely act on within defined guardrails, even in high-risk enterprise environments2

2

.

United Airlines Case Study: From 250 People to Proactive and Autonomous Management

Ramiro Zavala, Head of IT Operations, Observability and Quality at United Airlines, provided concrete evidence of this shift during his stage appearance with Dynatrace CMO Laura Heisman1

1

. Two years ago, diagnosing major incidents required upwards of 250 people. United operates more than 2,000 application services running continuously, where a single boarding pass triggers up to 500 unique services across mainframes, on-premises systems, and modern cloud applications.

Zavala's team consolidated fragmented monitoring tools onto Dynatrace, migrating approximately 800 applications in nine months through "developer Observability days" and "Dynatrace blitz months"1

1

. The results speak to the business impact: two best years on record consecutively, number one in the industry for on-time departures, and customer satisfaction scores up 2.6 points last year. Zavala credits the shift from signal-watching to business outcome focus, noting that "having a conversation about maybe a kiosk not printing the right number of bag tags is a lot different than a kiosk not being available"1

1

. The next phase involves agentic automation integrated with ServiceNow workflows.

Ecosystem of AI Agents and Enhanced Cloud Operations Capabilities

Dynatrace Intelligence includes an operator agent handling planning and orchestration, plus domain-specialized agents targeting SRE, development, and security teams1

1

. When high-impact vulnerabilities surface, the system assesses exposure automatically, correlates evidence across logs and traces, examines runtime behavior, and generates prioritized response plans flowing directly into ServiceNow, GitHub, Atlassian, AWS, Azure, Google Cloud, and other enterprise platforms1

1

2

2

.

The company also announced expanded cloud operations capabilities offering deeper cloud-native integrations across AWS, Azure, and Google Cloud, consolidating visibility into a single control plane2

2

. Additional updates include next-generation Real User Monitoring capabilities designed to unify frontend telemetry with backend context, addressing blind spots as organizations adopt AI-driven applications2

2

.

Channel Strategy Evolution Supports Partner-Led Lifecycle Approach

These announcements align with Dynatrace's four-year pivot to a channel-centric go-to-market strategy under CEO Rick McConnell3

3

. The company now works with approximately 700 channel partners, with about 80% of new sales bookings coming through and with the partner ecosystem last year, according to Jay Snyder, Senior Vice President of Partners and Alliances3

3

.

Snyder emphasized a shift "from the transaction to the lifecycle-based approach," where partners focus beyond initial deals to manage customers over time and deliver sustained success3

3

. Partners bring vertical industry expertise and help integrate Dynatrace technology with broader IT landscapes to automate remedial tasks and support cybersecurity systems. The company now provides partners with tools and frameworks from its own service organization, with some partners signing "teaming agreements" that spell out needs, deliverables, and expected outcomes3

3

. This approach positions organizations to move from reactive IT responses toward autonomous operations across digital ecosystems.