APM and Observability: Cutting Through the Confusion — Part 12

Pete Goldin
APMdigest

In Part 12, the final installment in the series, the experts present some final predictions about AI's future impact on APM and Observability.

Start with: APM and Observability - Cutting Through the Confusion - Part 11

AI-powered capabilities such as AI Assistants, zero-config ML-based multi-signal correlation, pattern analysis, failure detection, latency analysis, and more are enriching the APM experience and tightly integrating it with other observability signals, according to Bahubali Shetti, Senior Director, Product Marketing, Elastic. Users can solve problems holistically using all available signals and data, rather than relying on metrics, logs, or traces in isolation.

The integration of AI and machine learning will deepen, enabling faster, more accurate diagnostics and increasingly automated remediation, says Arun Balachandran, Senior Product Marketing Manager, ManageEngine APM Solutions.

"AI will become central, automating anomaly detection, root cause analysis, and performance optimization," adds Varma Kunaparaju, SVP and GM for Cloud Platform and OpsRamp Software, HPE, "making both APM and observability more proactive and predictive. This transformation will enable more agile and resilient IT operations, driving innovation and competitive advantage."

The following are more predictions from the experts:

ASSISTIVE OBSERVABILITY

Observability will move from being reactive to being assistive. As systems grow more complex, organizations will need observability platforms that don't just show what happened, but help explain why. That requires open, high-fidelity data, which is why the CNCF ecosystem is so critical. Projects like Thanos https://thanos.io/ (for scalable metrics), Fluent Bit (for log routing), and OpenTelemetry (for structured, correlated telemetry) are laying the foundation for AI-enhanced, team-centric observability that adapts as fast as the systems it observes.
Brian Douglas
Head of Ecosystem, Cloud Native Computing Foundation (CNCF)

AGENTIC WORKFLOWS AND CONVERSATIONAL EXPERIENCES

Agentic workflows and conversational experiences will completely change IT operations workflows, making it much more practical to find and resolve issues all through conversational experiences including code generation, patching, and deployment.
Bill Lobig
VP of Observability, IBM Automation

SMART DATA PIPELINES

Data pipelines will become smarter — filtering at the edge, routing to multiple destinations, and using AI to recommend what matters. 
Gurjeet Arora
CEO and Co-Founder, Observo AI

The implementation of AI engines is going to dramatically impact the scope and capabilities of APM solutions in a positive way. We will see the rise of tightly correlated data elements that are automatically traced, identified, and presented to IT Operations staff in real time, with targeted guidance on what they should be doing next to support the health of the application. Notably, this will be to a degree that will seem almost magical compared to current solutions. We're just seeing the very beginnings of it already, but I believe that vastly more data will be ingested and understood in real time, leading to what would be interpreted today as a near-perfect understanding of application state.
Bryan Cole
Director of Customer Engineering, Tricentis

OPENTELEMETRY

AI will clearly play a significant role in the future of observability, although it's unclear exactly which use cases will dominate. Some vendors today seem excited about the prospect of AI helping developers manage the overwhelming volume of data that comes from disparate logging, metrics, and APM tools. However, this data volume challenge isn't inherent to building systems — it's a consequence of emitting data in formats designed for previous generations of tooling. I hope that as tool makers bring AI into the observability landscape, they focus more on how AI can help us swiftly move into the OpenTelemetry future, e.g. by speeding the authoring and adoption of custom instrumentation, instead of providing an "intelligent" layer on top of a hodgepodge of existing logging, monitoring, and APM tooling. The higher up in the observability "funnel" we can deploy AI, the more powerful the results will be for our development teams.
Emily Nakashima
VP of Engineering, Honeycomb

SELF-LEARNING AI

In today's increasingly complex environments, visibility alone isn't enough. The next wave of AIOps solutions is being driven by self-learning AI platforms that unify and interpret data across operational domains, transforming it into predictive, prioritized, and actionable insights — without relying on static topologies or predefined rules. AIOps platforms built on a fully AI-native architecture are shifting the focus from simply monitoring systems to enabling intelligent, autonomous operations.

By applying predictive, causal, and generative AI, these platforms not only enhance the value of existing tools but increasingly have the potential to replace standalone observability solutions. They offer a single, intelligent layer that surfaces emerging issues, pinpoints root causes, and drives automated resolution — enabling a shift from fragmented monitoring to proactive, autonomous operations. Self-learning AI will ultimately replace traditional observability platforms by becoming the integrated, real-time source of operational truth. Rather than relying on topology and rules-based AIOps and Observability platforms, the next generation of platforms will continuously learn from live telemetry, historical incidents, human actions, and system behavior to proactively detect, diagnose, and even remediate issues. This real-time learning loop will reduce noise, surface meaningful patterns, and guide teams toward faster, more confident decisions — paving the way for predictive, autonomous, and eventually self-healing IT environments.
Josh Kindiger
President, Grokstream

SELF-HEALING SYSTEMS

It's likely we'll see more widely adopted "self-healing" application capabilities through the use of AI and the observability data that feeds it.
Justin Collier
Senior Director of Product Management, SmartBear

The next few years will be less about manually building SLAs, dashboards, and alerts and more centered on self-healing and adaptive systems. With the rise of AI and ML embedded into observability platforms, we'll see a shift toward systems that can detect anomalies, determine probable root causes, and even take corrective actions without (some) human intervention. The result is not just greater efficiency but a fundamentally more resilient and intelligent digital infrastructure.
Mimi Shalash
Observability Advisor at Splunk, a Cisco Company

We're heading toward Autonomous Service Reliability: systems that not only observe themselves, but also understand, diagnose, and even self-heal with minimal human intervention.
Severin Neumann
Head of Community & Developer Relations, Causely

AI WILL NOT REPLACE HUMANS

AI's role in APM and observability will be to assist and guide, not replace, human expertise. AI will be used to narrow down search spaces, prioritize issues, and support human operators in diagnosis rather than providing fully automated solutions.
Jeff Cobb
Global Head of Product & Design, Chronosphere

AI OBSERVABILITY

Expect to see a rise in the need for AI-specific observability and application performance monitoring. Customers embedding GenAI into their own offerings will need to monitor these AI applications and AI factories, driving new requirements for observability platforms that can handle these specialized workloads.
Paul Appleby
CEO, Virtana

Pete Goldin is Editor and Publisher of APMdigest

Related Links

Hot Topics

The Latest

Exploring AI Observability – Part 5: Roles

AI observability serves several roles across an organization. AI engineers use it to build, evaluate, and improve AI systems. SREs use it to operate those systems reliably and efficiently in production, including managing their cost. Leaders need it to connect AI behavior and operational performance to business outcomes. The common need is a clear understanding of whether AI systems are delivering the intended result ...

Exploring AI Observability – Part 4: Challenges

Part 4 outlines the challenges enterprise face when adopting AI observability. The main challenge is that AI complexity compounds at every stage of maturity. LLM-based applications introduce new behavior, dependencies, and quality considerations, while Agentic AI adds further failure modes as systems make decisions and take actions ...

Exploring AI Observability – Part 3: The Current State of AI Observability

Part 3 reports on the current state of AI observability. Most experts agree: enterprises are moving faster with AI than they are with the operational discipline needed to run it safely ...

Exploring AI Observability – Part 2: Traditional vs. AI Observability

When discussing AI Observability, one of the first questions that arises: Is this just an extension of traditional observability or is this a whole new type of tool? Part 2 compares traditional observability with AI observability ...

Exploring AI Observability – Part 1: Why It Matters

In this extensive 14-part series that will post over the next three weeks, experts across observability and AI will provide insight and opinions on the need for AI Observability, the challenges, the benefits, the current state, and predictions for the future ...

AgenticOps Scaling Quickly in the Enterprise

Amid rising complexity, organizations are moving quickly toward agent-driven operations: AgenticOps, according to The Impact of Agentic AI on Network Operations, a new report from Cisco ...

Why Better Technology Won't Fix a Broken Workflow

AI, automation, and enterprise software give organizations an extraordinary amount of leverage. Work that once required hours can happen in minutes. Information buried across systems can be analyzed almost instantly. Complex workflows can be coordinated across teams with far less manual intervention. That leverage creates a temptation to start with the technology ...

What If AI's Biggest Competitive Advantage Isn't the Model?

Everyone wants to talk about models. Compute clusters. Benchmark scores that climb a fraction of a point and make headlines anyway. It's an easy conversation to have because it's measurable. But it's also increasingly beside the point ...

Why Going All-In on Cloud No Longer Works in the Age of AI

For more than a decade, the default infrastructure play has been simple. Move more workloads to the cloud, retire on-premises gear and treat the data center as a cost to shrink. That logic held while public cloud offered near-limitless capacity at predictable prices, and AI demand is now testing both halves of that assumption ...

Visibility Is Not a Strategy - What Matters Is What Happens Next

Most enterprise IT leaders can tell you exactly how many applications are running across their organization. They can surface a dashboard showing telecom spend by region, pull a report on SaaS license utilization, or call up a real-time view of mobile device inventory. The data is there, and IT leaders have largely delivered on the promise of visibility. Yet, in most organizations, that visibility has not translated into business value ...

APM and Observability: Cutting Through the Confusion — Part 12

Pete Goldin
APMdigest

In Part 12, the final installment in the series, the experts present some final predictions about AI's future impact on APM and Observability.

Start with: APM and Observability - Cutting Through the Confusion - Part 11

AI-powered capabilities such as AI Assistants, zero-config ML-based multi-signal correlation, pattern analysis, failure detection, latency analysis, and more are enriching the APM experience and tightly integrating it with other observability signals, according to Bahubali Shetti, Senior Director, Product Marketing, Elastic. Users can solve problems holistically using all available signals and data, rather than relying on metrics, logs, or traces in isolation.

The integration of AI and machine learning will deepen, enabling faster, more accurate diagnostics and increasingly automated remediation, says Arun Balachandran, Senior Product Marketing Manager, ManageEngine APM Solutions.

"AI will become central, automating anomaly detection, root cause analysis, and performance optimization," adds Varma Kunaparaju, SVP and GM for Cloud Platform and OpsRamp Software, HPE, "making both APM and observability more proactive and predictive. This transformation will enable more agile and resilient IT operations, driving innovation and competitive advantage."

The following are more predictions from the experts:

ASSISTIVE OBSERVABILITY

Observability will move from being reactive to being assistive. As systems grow more complex, organizations will need observability platforms that don't just show what happened, but help explain why. That requires open, high-fidelity data, which is why the CNCF ecosystem is so critical. Projects like Thanos https://thanos.io/ (for scalable metrics), Fluent Bit (for log routing), and OpenTelemetry (for structured, correlated telemetry) are laying the foundation for AI-enhanced, team-centric observability that adapts as fast as the systems it observes.
Brian Douglas
Head of Ecosystem, Cloud Native Computing Foundation (CNCF)

AGENTIC WORKFLOWS AND CONVERSATIONAL EXPERIENCES

Agentic workflows and conversational experiences will completely change IT operations workflows, making it much more practical to find and resolve issues all through conversational experiences including code generation, patching, and deployment.
Bill Lobig
VP of Observability, IBM Automation

SMART DATA PIPELINES

Data pipelines will become smarter — filtering at the edge, routing to multiple destinations, and using AI to recommend what matters. 
Gurjeet Arora
CEO and Co-Founder, Observo AI

The implementation of AI engines is going to dramatically impact the scope and capabilities of APM solutions in a positive way. We will see the rise of tightly correlated data elements that are automatically traced, identified, and presented to IT Operations staff in real time, with targeted guidance on what they should be doing next to support the health of the application. Notably, this will be to a degree that will seem almost magical compared to current solutions. We're just seeing the very beginnings of it already, but I believe that vastly more data will be ingested and understood in real time, leading to what would be interpreted today as a near-perfect understanding of application state.
Bryan Cole
Director of Customer Engineering, Tricentis

OPENTELEMETRY

AI will clearly play a significant role in the future of observability, although it's unclear exactly which use cases will dominate. Some vendors today seem excited about the prospect of AI helping developers manage the overwhelming volume of data that comes from disparate logging, metrics, and APM tools. However, this data volume challenge isn't inherent to building systems — it's a consequence of emitting data in formats designed for previous generations of tooling. I hope that as tool makers bring AI into the observability landscape, they focus more on how AI can help us swiftly move into the OpenTelemetry future, e.g. by speeding the authoring and adoption of custom instrumentation, instead of providing an "intelligent" layer on top of a hodgepodge of existing logging, monitoring, and APM tooling. The higher up in the observability "funnel" we can deploy AI, the more powerful the results will be for our development teams.
Emily Nakashima
VP of Engineering, Honeycomb

SELF-LEARNING AI

In today's increasingly complex environments, visibility alone isn't enough. The next wave of AIOps solutions is being driven by self-learning AI platforms that unify and interpret data across operational domains, transforming it into predictive, prioritized, and actionable insights — without relying on static topologies or predefined rules. AIOps platforms built on a fully AI-native architecture are shifting the focus from simply monitoring systems to enabling intelligent, autonomous operations.

By applying predictive, causal, and generative AI, these platforms not only enhance the value of existing tools but increasingly have the potential to replace standalone observability solutions. They offer a single, intelligent layer that surfaces emerging issues, pinpoints root causes, and drives automated resolution — enabling a shift from fragmented monitoring to proactive, autonomous operations. Self-learning AI will ultimately replace traditional observability platforms by becoming the integrated, real-time source of operational truth. Rather than relying on topology and rules-based AIOps and Observability platforms, the next generation of platforms will continuously learn from live telemetry, historical incidents, human actions, and system behavior to proactively detect, diagnose, and even remediate issues. This real-time learning loop will reduce noise, surface meaningful patterns, and guide teams toward faster, more confident decisions — paving the way for predictive, autonomous, and eventually self-healing IT environments.
Josh Kindiger
President, Grokstream

SELF-HEALING SYSTEMS

It's likely we'll see more widely adopted "self-healing" application capabilities through the use of AI and the observability data that feeds it.
Justin Collier
Senior Director of Product Management, SmartBear

The next few years will be less about manually building SLAs, dashboards, and alerts and more centered on self-healing and adaptive systems. With the rise of AI and ML embedded into observability platforms, we'll see a shift toward systems that can detect anomalies, determine probable root causes, and even take corrective actions without (some) human intervention. The result is not just greater efficiency but a fundamentally more resilient and intelligent digital infrastructure.
Mimi Shalash
Observability Advisor at Splunk, a Cisco Company

We're heading toward Autonomous Service Reliability: systems that not only observe themselves, but also understand, diagnose, and even self-heal with minimal human intervention.
Severin Neumann
Head of Community & Developer Relations, Causely

AI WILL NOT REPLACE HUMANS

AI's role in APM and observability will be to assist and guide, not replace, human expertise. AI will be used to narrow down search spaces, prioritize issues, and support human operators in diagnosis rather than providing fully automated solutions.
Jeff Cobb
Global Head of Product & Design, Chronosphere

AI OBSERVABILITY

Expect to see a rise in the need for AI-specific observability and application performance monitoring. Customers embedding GenAI into their own offerings will need to monitor these AI applications and AI factories, driving new requirements for observability platforms that can handle these specialized workloads.
Paul Appleby
CEO, Virtana

Pete Goldin is Editor and Publisher of APMdigest

Related Links

Hot Topics

The Latest

Exploring AI Observability – Part 5: Roles

AI observability serves several roles across an organization. AI engineers use it to build, evaluate, and improve AI systems. SREs use it to operate those systems reliably and efficiently in production, including managing their cost. Leaders need it to connect AI behavior and operational performance to business outcomes. The common need is a clear understanding of whether AI systems are delivering the intended result ...

Exploring AI Observability – Part 4: Challenges

Part 4 outlines the challenges enterprise face when adopting AI observability. The main challenge is that AI complexity compounds at every stage of maturity. LLM-based applications introduce new behavior, dependencies, and quality considerations, while Agentic AI adds further failure modes as systems make decisions and take actions ...

Exploring AI Observability – Part 3: The Current State of AI Observability

Part 3 reports on the current state of AI observability. Most experts agree: enterprises are moving faster with AI than they are with the operational discipline needed to run it safely ...

Exploring AI Observability – Part 2: Traditional vs. AI Observability

When discussing AI Observability, one of the first questions that arises: Is this just an extension of traditional observability or is this a whole new type of tool? Part 2 compares traditional observability with AI observability ...

Exploring AI Observability – Part 1: Why It Matters

In this extensive 14-part series that will post over the next three weeks, experts across observability and AI will provide insight and opinions on the need for AI Observability, the challenges, the benefits, the current state, and predictions for the future ...

AgenticOps Scaling Quickly in the Enterprise

Amid rising complexity, organizations are moving quickly toward agent-driven operations: AgenticOps, according to The Impact of Agentic AI on Network Operations, a new report from Cisco ...

Why Better Technology Won't Fix a Broken Workflow

AI, automation, and enterprise software give organizations an extraordinary amount of leverage. Work that once required hours can happen in minutes. Information buried across systems can be analyzed almost instantly. Complex workflows can be coordinated across teams with far less manual intervention. That leverage creates a temptation to start with the technology ...

What If AI's Biggest Competitive Advantage Isn't the Model?

Everyone wants to talk about models. Compute clusters. Benchmark scores that climb a fraction of a point and make headlines anyway. It's an easy conversation to have because it's measurable. But it's also increasingly beside the point ...

Why Going All-In on Cloud No Longer Works in the Age of AI

For more than a decade, the default infrastructure play has been simple. Move more workloads to the cloud, retire on-premises gear and treat the data center as a cost to shrink. That logic held while public cloud offered near-limitless capacity at predictable prices, and AI demand is now testing both halves of that assumption ...

Visibility Is Not a Strategy - What Matters Is What Happens Next

Most enterprise IT leaders can tell you exactly how many applications are running across their organization. They can surface a dashboard showing telecom spend by region, pull a report on SaaS license utilization, or call up a real-time view of mobile device inventory. The data is there, and IT leaders have largely delivered on the promise of visibility. Yet, in most organizations, that visibility has not translated into business value ...