Dynatrace (DT): 800+ Customers Run Operations Autonomously With Agentic AI

On its 2026-08-05 earnings call, Dynatrace, Inc. (DT) said more than 800 customers are running operations autonomously with its agentic AI capabilities, up from roughly 500 the previous quarter. Management tied those agent actions directly to usage-based platform consumption, but has not disclosed how much revenue they contribute[1].

What Dynatrace sells and where the AI sits

Dynatrace (DT) sells a software observability platform to large enterprises, which connect their application, cloud infrastructure and log data and pay by subscription. The usage-based Dynatrace Platform Subscription (DPS) is the main contract form: the more a customer uses, the more it consumes.

The application is the AI operations capability built into that platform. Operations, development and security teams connect production logs, metrics and traces; the AI finds root causes, predicts likely problems, proposes fixes, and increasingly lets operations agents execute the fix directly. The company calls this combination of deterministic and agentic AI "Dynatrace Intelligence." It is embedded in the platform, open to all customers and not priced separately[2]. On 2026-08-05 the company also launched Blue Box, which exposes the live state of production environments to developers and to the AI agents that write code for them; once live, agents diagnose the cause and return a fix[1]. The capability sits in the company's core business.

From analysis engine to autonomous remediation

The disclosure trail starts with the 2023-11-02 earnings call. Management said customers had used Davis's causal and predictive AI for more than a decade, that it gave operations teams precise answers and drove automation, and that the generative Davis Copilot was still in development with no monetization decided[3]. At that point the application produced answers but did not act on them.

The first turning point came on 2026-02-09. Deterministic and agentic AI were merged into Dynatrace Intelligence, described as an agentic operations system, and connected to the agent ecosystems of Amazon, Azure and GCP. Management acknowledged that autonomous prevention, remediation and optimization were still a direction being realized, and that agents could take over only part of the actions[2].

The second, on 2026-05-13, made it a production feature with a countable customer base. Agents across operations, development and security were delivered and were coordinating end-to-end actions in customer production environments; more than 500 customers were deploying them to run operations autonomously and connecting them to AI development tools such as GitHub Copilot[4].

By 2026-08-05 the count exceeded 800, and Blue Box extended the reach to developers and coding agents. The analysis engine had reached the stage of executing fixes on its own[1].

What the 800-customer figure does and does not measure

The 800 figure counts customers running operations autonomously with the company's agentic capabilities, up from roughly 500 the prior quarter. The company has not defined the threshold for "autonomous," has not disclosed usage depth, and has not split the count across the operations, development and security domains[1].

Customer-side results exist only as individual examples. Canadian telecom company Telus used AI to cut its average time to resolve issues from forty minutes to five minutes; that is a whole-company result reported by the customer, with incident scope, measurement period and the share attributable to AI alone undisclosed[2]. The 2026-08-05 call also cited a customer that used Dynatrace as its system of record while developing a CRM system with AI assistance and reported roughly seven-figure savings. The currency, amount and period were not specified, the savings came from the customer's own AI development project, and Dynatrace's share was not separated[1]. These figures rest on different bases and cannot be added together.

How agent actions feed subscription revenue and ARR

The financial metrics this application maps to are subscription revenue and annual recurring revenue (ARR), and the entry point is platform consumption by usage-based (DPS) customers. On 2026-08-05 management described the mechanism directly for the first time: every time a customer asks Dynatrace Intelligence for an answer through an AI function call or an MCP integration, or an SRE agent autonomously executes a remediation action, the event counts toward DPS usage. More customers running operations autonomously and more frequent agent actions mean higher consumption, which flows into subscription revenue and ARR through overages and expanded renewals[1]. This is consumption earned by selling AI capability, not savings from the company using AI on its own costs.

The revenue contribution remains unquantified. Management's only comparison is that consumption growth among AI customer cohorts is 1.5 times that of non-AI cohorts. That comparison combines AI observability customers with agentic customers, states no measurement period, and the company has not broken out the consumption, revenue or ARR contributed by agentic operations alone[1].

What can be confirmed so far

Agentic operations has moved from a roadmap item to a feature that hundreds of customers use in production, positioned as an entry point for usage-based consumption. How much consumption, revenue or ARR those agent actions generate has not been quantified separately. Only when the company discloses the consumption growth of its autonomous-operations cohort, or the dollar contribution of agent actions to DPS consumption revenue and ARR, separately from AI observability customers, will this path's effect on revenue growth become a checkable figure rather than a management statement.

Application assessment

• Dynatrace Autonomous Operations | Business position: core business | Deployment stage: limited production | Coverage: multiple businesses or regions | Value type: revenue growth

Sources

[1] Drillr · Dynatrace, Inc. (DT) · 2026-08-05 · Earnings call

Quote: And more than 800 are running operations autonomously with Dynatrace's agentic capabilities, up from roughly 500 last quarter. Additionally, consumption growth for customers in these AI cohorts is 1.5 times higher than that of non-AI cohort customers.

[2] Drillr · Dynatrace, Inc. (DT) · 2026-02-09 · Earnings call

Quote: Canadian communications giant Telus shared how they are using AI to move from firefighting to proactive reliability, reducing the average time to resolve issues from forty minutes to five minutes.

[3] Drillr · Dynatrace, Inc. (DT) · 2023-11-02 · Earnings call

Quote: our customers have benefited from Davis for more than a decade. Last quarter, I shared airplanes to add a third critical element to our existing Davis AI architecture, a generative AI capability named Davis Copilot.

[4] Drillr · Dynatrace, Inc. (DT) · 2026-05-13 · Earnings call

Quote: More than 500 customers are deploying a Dynatrace's agentic capabilities to run operations autonomously and extend that intelligence into AI development tools like QuadCode and GitHub Copilot.