Using AI to Reimagine Business Operations
Eric Johnson doesn’t frame AI as hype or a futuristic buzzword. For the CIO of PagerDuty, a leading digital operations management platform, artificial intelligence and automation are already transforming how enterprises respond to incidents, streamline operations, and scale efficiency. In his view, AI is not simply a nice-to-have tool; it is a core lever for business impact and operational resilience.
He describes PagerDuty’s evolution in terms more often associated with enterprise data strategy and operational engineering than traditional IT support. It’s a signal of how AI is fusing disciplines and turning reactive incident management into an intelligent, proactive operations engine.
Why Legacy Channels Don’t Cut It
PagerDuty began as a notification technology: alerting teams when something went wrong and notifying who was on call. As Johnson explains:
“The company has evolved quite a bit, from not just being able to tell you there's a problem and alerting you to a problem, but getting into a place where it could even see that there's a problem, understand what the problem is, and potentially have a workflow that you could build to help you resolve the problem.”
With cloud infrastructure, distributed teams, remote operations, and complex dependencies, simply alerting isn’t enough. Outages or failures often emerge from complicated interactions across systems, infrastructure, and data pipelines. As Johnson says:
“Maybe you got infrastructure or some data integration that looks like it's not working like it should. There are early indications you could have a problem and be able to fix the problem before it causes an outage.”
This transformation is essential. The environments enterprises run today are far more complex than traditional on-premise servers with blinking lights. With cloud services, remote work, hybrid infrastructure, and services interconnected globally, organizations need tools that can scale, correlate data across layers, and operate in real time. PagerDuty’s repositioning reflects exactly that shift.
Prioritizing Data Strategy
Before even considering AI or automation, Johnson insists that enterprises must get their data strategy right. He frames it as a prerequisite for meaningful AI adoption: “You’ve got to first think through how you are using data to drive business impact. What are very clear use cases that you could be focusing on around generative AI?”
He warns against a common pitfall: companies chasing the “bright, shiny AI object,” then searching for a problem to match it, rather than first identifying real business problems and then looking to AI (or automation) for solutions.
Moreover, Johnson emphasizes the importance of data quality, governance, and trust: “If your fundamentals aren’t in good shape, it’s going to be really hard to take advantage of some of this really, really cutting-edge, innovative technology. Data is going to be something that CIOs are going to have to get really maniacal about and super focused on making sure that it is world class.”
In other words: AI isn’t magic. It’s only as effective as the data and structures behind it.
From Alerts to Automated Remediation
What excites Johnson most is what PagerDuty is building now, a shift from being a passive alerting system to an active, intelligent operations platform, blending AI, automation, and workflow orchestration.
He outlines a multi-phase evolution:
- Phase 1: Detection & diagnosis: Use AI to sift through alerts, log data, telemetry, and signals, then identify not just that “something is wrong,” but what is wrong and where.
- Phase 2: Automated response & remediation: Once the issue is identified, a built-in workflow can trigger the remediation (e.g., rerouting traffic, restarting servers, rolling back deployments, or alerting appropriate teams), reducing human friction and response time.
- Phase 3: Predictive prevention using AI + pattern recognition: With enough data, AI could detect early signals of potential failures like anomalies, abnormal resource usage, emerging patterns, and proactively act or flag to avoid outages before they occur. Johnson sees this as the long-term vision.
As Johnson puts it: “We believe using AI to be able to identify all of these data points helps our customers go from ‘I’ve identified a problem’ to ‘I’ve actually remediated and closed a problem’ much, much faster.”
This is AI-augmented operations, where detection, diagnosis, and even remediation can flow without manual handoffs, or at least with minimal human intervention.
Using AI to Drive Business Impact
Johnson doesn’t view AI as a purely technical play. For him, AI + automation are business levers, tools to help organizations scale without scaling headcount, especially now that growth emphasizes profitability and operational discipline over “growth at any cost.”
He suggests starting small but strategic, particularly with functions like sales, marketing, customer success, or finance: “You know, what are some of the key business drivers that they have, how are they measuring things like churn, expansion. Having an understanding and then building data products to support those things and move the needle.”
In essence: treat AI as a tool to deliver business metrics, not just technical KPIs. Whether it’s optimizing conversion, reducing customer churn, improving time to resolution, or cutting operating costs, AI should be aligned with measurable outcomes.
Johnson warns that many IT organizations are good at delivering features, but poor at measuring and positioning their true impact. AI and automation, if deployed purposefully, can change that narrative.
Recognizing Both Promise and Limits
Despite his enthusiasm, Johnson is realistic about the limitations of generative AI and automation, especially around legacy systems. He argues that while generative AI and “copilot-style” tools are powerful, they are not a panacea: “I don’t know how great AI is at taking a bunch of old legacy code and untangling that.”
In his experience, most of the world’s code is legacy code. It often requires domain knowledge, context, and human judgment, especially when security, compliance, or deep integration is involved. As he puts it: “you still need a human to kind of go through and untangle that many times.”
He also stresses the importance of change management. Automation doesn’t just affect systems. It affects people, roles, workflows, and culture. Moving to a future with bots, virtual agents, and AI-assisted processes means shifting mindsets and reskilling. “Your job’s going to be different. There’s going to be new skills you’re going to develop.”
Without that human-centered approach, AI transformation risks stalling, or worse, creating resistance and friction inside the organization.
AI as the New Operations Playbook
Eric Johnson’s vision is built on practical experience, a deep understanding of data fundamentals, and a clear-eyed view of what automation can realistically achieve.
Through PagerDuty, he is building what many enterprises will need in the next wave: a system that not only detects issues in complex, cloud-driven environments, but diagnoses root causes, automates remediation, and even predicts future problems before they occur.
AI and automation, in this view, are not just tools; they are the new operations playbook. And the CIO’s job is about keeping systems alive and steering the business, aligning technology with commercial impact, and building teams capable of delivering that impact.