Dynatrace Frames $915M Arize Deal Around AI Agents Taking Action

A new discussion of the proposed acquisition shifts the emphasis from dashboards to the data AI agents may use to diagnose problems and recommend or initiate fixes.

By 2 min read
Dynatrace Frames $915M Arize Deal Around AI Agents Taking Action
Dynatrace Frames $915M Arize Deal Around AI Agents Taking Action

Listen to this story

The audio brief

About 1:40
0:001:40
Read transcript
Dynatrace is framing its proposed nine-hundred-fifteen-million-dollar purchase of Arize around a question that comes after an alert: can an AI agent use observability data to diagnose a problem, recommend a fix, or eventually start the remediation itself? The deal would combine two different views of an AI system. Arize records what happens inside an AI application, including model calls, retrieval, and tool use. Dynatrace tracks telemetry across the applications, services, and infrastructure that application depends on. Together, the companies say those records could help teams follow an AI failure from the agent’s behavior into the broader software stack. That matters because AI systems can produce different answers from similar inputs. Checking whether a service is online is not enough; teams also need to evaluate the quality of the response. Arize co-founder Aparna Dhinakaran described evaluation as more than a right-or-wrong test. The intended next step is to turn this shared context into recommendations or automated action for developers, AI engineers, and site-reliability teams. But that is an integration goal, not a product available today. The acquisition still needs regulatory and customary approvals, with closing expected later this quarter or early in Dynatrace’s fiscal third quarter. Dynatrace forecasts about two hundred basis points of fiscal 2027 ARR-growth benefit, alongside a roughly one-hundred-seventy-five-basis-point reduction in non-GAAP operating margin. The key constraint is whether the two systems can be integrated well enough for agents to act on trustworthy, connected records.

Story brief

3 key points

Dynatrace’s proposed $915 million acquisition of Arize is positioned as a move toward observability systems that can inform—and eventually trigger—AI-agent remediation. Arize contributes traces of model calls, retrieval, and tool use; Dynatrace adds telemetry across applications, services, and infrastructure. The deal remains pending regulatory and customary approvals, with closing expected later this quarter or...

  1. 01

    The transaction includes about $815 million in cash plus replacement equity awards for Arize employees.

  2. 02

    Arize founders Jason Lopatecki and Aparna Dhinakaran are expected to join Dynatrace; Lopatecki would lead the Arize team.

  3. 03

    The combined view is intended to trace AI failures through the broader software stack, including response quality and service health.

The case for Dynatrace’s proposed purchase of Arize is increasingly about what happens after an alert. In a September discussion of the deal, the companies described a model in which AI agents consume observability data to identify problems, recommend changes or initiate remediation. Dynatrace’s earlier $915 million agreement to acquire Arize would bring together the records those agents need: AI behavior on one side, and application and infrastructure telemetry on the other.

The transaction has not closed. Dynatrace said it expects closing later this quarter or early in its fiscal third quarter, subject to regulatory reviews and other customary conditions. Arize founders Jason Lopatecki and Aparna Dhinakaran are expected to join Dynatrace at closing, with Lopatecki continuing to lead the Arize team.

Linking an agent’s path to the systems it touches

Arize builds tools to trace, evaluate and improve AI applications and agents. Its traces record model calls, retrieval and tool use during a run. Dynatrace correlates telemetry and traces dependencies across applications, services and infrastructure. Dynatrace says joining those views is intended to help teams trace AI-related failures through the broader software stack.

That connection addresses a different problem from simply checking whether software is online. AI systems can produce varying outputs from similar inputs, and teams must assess response quality as well as the health of the services around them. Dhinakaran said evaluation is no longer only a right-or-wrong test, but a way to measure the quality of responses.

From a shared view to automated action

Dynatrace says the combined offering would cover evaluating, operating and continuously improving AI applications from development through production. Its wider AI and automation strategy already includes Dynatrace Intelligence and BlueBox AI for agentic development and site-reliability engineering workflows.

The appeal is a shared context for developers, site-reliability engineers, platform teams, AI engineers and data scientists. But it also raises the standard for the underlying records: data used to guide an agent’s recommendation or remediation needs to connect its actions with the surrounding software environment. The companies describe that capability as an intended result of integration, rather than a product available today.

A deal with a financial and integration hurdle

The transaction consists of roughly $815 million in cash plus replacement equity awards for Arize employees. Dynatrace expects it to add about 200 basis points to fiscal 2027 annual recurring revenue growth and reduce non-GAAP operating margin by about 175 basis points. Those are company forecasts, and the larger promise still depends on closing and integrating two systems built for different parts of the AI operations problem.

Sources

  1. dynatrace.comDynatrace to Acquire AI Observability Leader Arize
  2. arize.comArize and Dynatrace: Making the World's AI Work
  3. siliconangle.comObservability shifts as AI agents reshape operations - SiliconANGLE
  4. devops.comDynatrace Acquires Arize as AI Agents Deepen the Observability Challenge - DevOps.com

Loading discussion...