Dynatrace has completed its acquisition of Arize, bringing AI observability tools for tracing, evaluation, and experimentation into the company’s broader application and infrastructure monitoring portfolio.

The acquisition connects two sets of engineering workflows: Arize’s tools for examining AI applications and agents, and Dynatrace’s observability platform for SRE and platform engineering teams. Dynatrace describes the combination as a foundation for observability across AI development, deployment, operation, and improvement. Product integration remains work ahead.

Steve Tack, Chief Product Officer at Dynatrace, explained: “Technical health alone does not determine whether an AI system is working as intended. Teams need context across the lifecycle to understand what happened, identify where an issue originated, assess what it affected, and apply what they learn to the next iteration.”

Dynatrace announced its intent to acquire Arize in August. With the transaction complete, the companies are beginning to develop a joint product and platform roadmap, with priorities to be informed by customers and developers.

Arize traces agent behaviour and outcomes

Arize’s workflows let AI engineers inspect agent trajectories, including model calls, retrieved context, tool use, performance, and cost. Teams can run evaluations, compare experiments, and use the results to guide subsequent changes to an application or agent.

Dynatrace provides context across the surrounding applications, services, infrastructure, user experiences, and business processes. The proposed integration would connect those operational records with the evidence AI engineers use to assess behaviour.

Tack’s explanation separates service availability from the correctness of an AI system’s output. An application can return a plausible but incorrect response without triggering an error. An agent can finish a workflow while using stale context or updating the wrong record, even when the supporting services appear healthy.

Tracing the path through models, retrieved information, and tools provides evidence for investigating those failures. Tack argues that application and infrastructure telemetry alone cannot explain an agent’s actions or the cost of its trajectory.

Changes to models, prompts, retrieval strategies, tools, data, and agent workflows are identified as reasons to continue evaluation after deployment. Findings from investigations can become datasets, regression tests, evaluators, or experiments for testing the next iteration. Connecting those activities with production context is part of Dynatrace’s stated plan, rather than a completed integration.

Stephen Elliot, IDC Group VP for Software Development and IT Operations, commented: “As agents multiply across enterprises, the importance of observability has risen. Bringing evaluation and observability together closes the loop between building AI applications and running them reliably in production, giving teams the visibility they need to catch issues earlier and resolve them faster.”

Phoenix and OpenInference retain open-source roles

Phoenix supports tracing, evaluation, failure investigation, dataset curation, and experiment comparison as an open-source project. Arize AX provides these workflows through a managed platform spanning development and production.

Phoenix is built on OpenTelemetry and OpenInference. Arize created OpenInference as an open-source instrumentation project and set of semantic conventions for AI observability.

OpenTelemetry formally accepted a code grant of OpenInference GenAI instrumentation from Arize in June 2026. That code is being incorporated incrementally into OpenTelemetry’s GenAI instrumentation project. OpenInference continues as an open, OpenTelemetry-compatible project.

Dynatrace plans to support the open-source approach while examining how to connect Phoenix and Arize AX workflows with its platform’s operational context. Tack said the near-term product work would identify where those connections could provide value across the AI lifecycle.

Arize CEO Jason Lopatecki, CPO Aparna Dhinakaran, and the Arize team are joining Dynatrace. Arize AX and the Dynatrace platform remain available as standalone offerings, and Phoenix remains available as an open-source project.

See also: Atlassian moves 100,000-host metrics platform to OpenTelemetry

Want to learn more about cybersecurity from industry leaders? Check out Cyber Security & Cloud Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the AI & Big Data Expo. Click here for more information.

Developer is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.