Disclaimer : This article reflects our personal perspectives and experiences and does not represent the views of our employers or ServiceNow. We have used Generative AI as a supporting tool for selected visuals, concept exploration, and content refinement. The experiences, architecture thinking, observations, and conclusions shared here are our own.

 

Lets continue from our earlier discussion ... If you missed my first part , link is here  ; part 2 , here 

 

_ Zero-Touch Doesn’t Mean Zero People _

One question almost always comes up when we talk about Zero-Touch IT: “If operations become autonomous, what happens to the people?”

From our experience, Zero-Touch is not about removing people from IT Operations. It is about removing people from predictable, repetitive operational work.

Think about how much engineering time is still spent collecting the same diagnostic information, checking dashboards, reviewing logs, correlating alerts, restarting known services, validating recovery, and documenting what happened. If the environment already understands the issue, knows the approved remediation and has enough confidence to execute it safely, requiring an engineer to repeat those steps every time adds very little value.

That is where the human role starts to change. Instead of engineers constantly executing known procedures, they increasingly define the conditions under which automation is allowed to operate. They establish guardrails, approval thresholds, exception paths and recovery mechanisms. And when the situation is unfamiliar, risky or ambiguous, the platform should know when to stop and bring a person back into the loop.

AI Agents can make this even more powerful. They can help bring together alerts, CMDB and service context, recent changes, historical incidents and knowledge before recommending—or, where appropriate, executing—the next action. But an AI Agent is only as effective as the operational foundation around it. Poor CMDB data, missing service relationships, weak knowledge or unreliable automation do not suddenly disappear because AI is introduced.

The Role of People Is Changing, Not Disappearing

For us, the progression looks something like this:

Human executes → Human approves → Human handles exceptions → Human governs and improves.

As confidence grows, human effort moves away from repetitive execution and toward engineering, governance, exception management and continuous improvement. And this is an important distinction: the objective is not maximum automation. The objective is the right level of automation for the risk, context and business impact involved.

Key Takeaway :

Zero-Touch IT doesn’t remove people from operations. It removes unnecessary touches and allows people to focus on decisions, exceptions and improvements that actually require human judgment.

That sounds compelling in theory. But there is a much harder question we hear from organizations:

“Where do we actually start?”

Trying to make the entire IT Operations organization autonomous is rarely the right first move. The better approach is to identify the right use cases, prove them safely, measure the outcome and expand from there.

That is where Part 4 begins.