The IT Professional Is Not Becoming Obsolete. The Job Description Is.

Author: Chetan Saundankar | Founder and CEO of Coditation

Every IT Professionals Day now comes with the same undercurrent: appreciation for the people who keep the lights on, shadowed by a quiet question about whether AI is about to turn the lights off on their careers.

That question is based on a false premise. AI isn’t competing with IT professionals for the same job. It is changing what the job is, moving value away from execution and toward judgment, accountability, and decision-making.

Learn more: Write for Us + Technology

By “IT professional,” I don’t mean anyone whose title happens to contain the letters IT. I mean the people responsible for turning technology into a working business system, the ones who decide what a system should do, not just keep it running. That distinction is the whole argument.

For twenty years, that role was measured by execution: how fast you could patch, migrate, build, and resolve. Execution was scarce, so execution was what got hired, trained, and promoted. AI doesn’t destroy the skills IT professionals build careers on. It destroys their scarcity. And when a skill stops being scarce, the value attached to it doesn’t disappear, it moves to whatever’s scarce now: judgment, and the willingness to be accountable for it.

That shift shows up in three places.

The first is technology selection. Somewhere in the last two years, “where can we use AI” quietly replaced “what’s the simplest technology that solves this reliably.” Those aren’t the same question, and conflating them is expensive. What if the most valuable IT professional right now isn’t the one deploying AI fastest, but the one willing to say it’s the wrong tool for this particular problem?

Some of the strongest engineering outcomes I’ve seen came from teams that treated a large language model as one option among several, and chose plain, well-understood optimization instead, because that’s what the problem actually called for. Knowing the shape of a problem well enough to pick the right-sized tool for it, instead of the fashionable one, is a form of judgment that becomes more valuable, not less, as capable technology becomes abundant.

The second is who holds the authority when AI is doing the work. The common assumption is that a human in the loop is a temporary safeguard, removed once the system proves itself. In practice, the opposite tends to happen: systems earn trust when a person retains real authority over what the AI decides, not just visibility into it after the fact.

That’s a meaningful distinction, there’s a difference between a human reviewing an AI’s output and a human owning the decision the AI is helping make. The goal was never to keep people doing AI’s work by hand. It’s to make sure someone remains answerable for the decision, which is a different job than the one being automated.

The third is becoming the translator between institutional knowledge and new systems. In one engagement, a client was running fifteen-plus legacy applications, years old, with the reasoning behind key decisions long undocumented. AI agents were used to work through that history, and engineers used the output to guide a modernization that had originally been estimated at eight and a half months, completed in four.

The time saved wasn’t the interesting part. What changed was the engineers’ role: instead of spending months personally deciphering old code, they spent that time validating what the AI surfaced, correcting it where it was wrong, and deciding what should and shouldn’t carry forward into the new system. AI didn’t replace their expertise. It gave them a faster way to apply it.

Put together, these describe a different job than the one most IT organizations still hire and evaluate people against. Diagnose the problem. Decide what should be automated and what shouldn’t. Orchestrate the systems doing the work. Verify what they produce. Own the outcome, regardless of what actually generated it. Execution is increasingly becoming the easier part of that chain. The harder part, the part that was always harder to measure, is the judgment behind it.

The IT professional of the AI era isn’t the person standing between a business and its technology. It’s the person deciding what the technology should be allowed to do in the first place. AI may take more execution out of the job. It won’t take the responsibility out of it. If anything, it makes responsibility the job.

Contributor: Chetan Saundankar is the Founder and CEO of Coditation, a data, AI and product engineering company helping healthcare organisations deploy AI to improve operational efficiency. He is also the founder of Plant360.ai, an AI platform for industrial engineering and operations and has 20+ years of experience in the field.