The inherently composable nature of modern enterprise software stacks grants architectural freedom. Software developers leverage componentized and containerized logic to create optimized code deployments, which can fluidly shift between workloads across multi-cloud instances.
Agentic functions are now enjoying similar freedom of movement, yet a lack of standardized AI agent telemetry leaves us in a “Wild West” scenario.
Developers are empowering production agents with the ability to call multiple system tools, invoke connections to diverse AI models (language, visual, large, and small), and even “improve” user requests, handing off work to other domain-specific agents.
While this is excellent news for system adaptability, it represents a looming nightmare for agent telemetry.
As Richard Young of Arize states, “When you use standards like OpenTelemetry and OpenInference, you keep optionality without losing visibility… The trace format stays consistent even as the stack changes.”
Why Agent Telemetry Matters
Within the broader universe of observability, telemetry at this level provides software engineers with crucial insights into where agents exist, what connections they are authorized to use, and what actions they have taken.
Richard Young, Technical Director of Partner Solutions Architecture at AI agent engineering company Arize, emphasizes that the agent telemetry challenge isn't about the existence of integration points. He asserts that the “important part is portability,” not just of agents, but of the telemetry standards used to measure them.
Young elaborates on his organization’s blog: “When you use standards like OpenTelemetry and OpenInference, you keep optionality without losing visibility. Standardized agent telemetry lets you change frameworks, models, tools, or observability backends without rebuilding your instrumentation every time. The trace format stays consistent even as the stack changes.”
For Young, the critical narrative isn't about point-to-point integrations, but the drive towards a shared telemetry model for agents.
Google Cloud & Arize AI Partnership
Arize is partnering with Google Cloud subsequent to the hyperscaler launching its Gemini Enterprise Agent Platform last month. The Arize AX enterprise agent development platform not only receives traces (chronological records of software execution event history) from the Gemini Agent service, but it also aligns agent telemetry around OpenTelemetry and OpenInference. This strategy ensures software engineering teams can instrument agents once, analyze behavior consistently, and avoid locking critical observability data inside a single platform.
Ryan Mangan, CEO of cloud resource optimization company EfficientEther, tells The New Stack that in any live production software deployment, “you can’t operate what you can’t see,” a principle doubly true for complex agentic workflows.
A single agent run can involve intricate steps such as request rewriting, retrieval, multiple tool and model calls, retries, and handoffs, necessitating robust, standardized telemetry for effective management and debugging.