eccher.
PT
Management

HI-C is not a career trend.

· · 6 min read
Translated from Portuguese. Read the original

The HI-C thesis is going viral. HI-C stands for High-Impact Individual Contributor, which is what Elena Verna called someone who single-handedly delivers a project with business value, end to end. In tech, that's the senior engineer who uses AI to deliver entire projects alone.

The narrative features former leaders (managers, directors, VPs) going back to IC roles because they can now do what used to take a squad. It's sold as the new career flex: someone with upper-management responsibility and pay, but no direct reports. On social media, it also shows up as the engineer's victory over management, a revenge story, the end of the pod era.

I don't read it that way. To me, what's happening is structural more than it is about careers: the unit of production in tech teams is shifting away from the squad, and it isn't clear yet what takes its place.

What I've been seeing in practice

I've been watching this from the outside, through the talk that has taken over social media and in conversations with a few people. Senior people who get put on a single project, with Claude or Cursor at hand, run everything end to end: discovery, code, deploy, communication. They deliver in weeks what used to take a squad months.

I'm describing something I've observed, without the content-creator hyperbole. It repeats often enough that I've stopped treating it as an exception, though not yet often enough to call it a rule. That middle zone interests me more than the certainty of the people selling the model or the skepticism of the people who treat it as passing hype.

Why the org chart gets stuck

The structure of the teams I know, mine included, was designed around an unspoken assumption: the smallest unit of delivery is a group. Since one person alone can't handle a product's complexity end to end, teams get built as pods, pairs, and a hierarchy of coordinator, lead, and dev. When a senior person shows up who could deliver an entire project alone, the manager slots them in as a pod resource, because that's the only way the org chart knows how to read them. Or the manager carves out a one-off exception to fit them in.

When that senior person asks to leave the leadership track and go back to working on solo projects, the system reads it as a step back. Their salary growth stalls, they drop out of succession plans, and within a few months the most productive person on the team is already thinking about leaving. Management still hasn't learned how to invest in this profile, and the engineer isn't rejecting management.

I'll admit that at first I thought all of this was just another career fad, the kind that passes once the AI hype cools. The problem with that view is that it treats a structural phenomenon as a cultural preference. Whether or not the engineer wants to go back to being an IC matters much less than the fact that there's a class of work that used to need six people and that today, in some cases, one person can handle. The question stops being about careers and becomes about how the manager puts the team together. On that, I don't have the answer, and I think we should be honest about it.

What's missing before any change: metrics

At the next budget review, someone from finance is going to ask what this person delivers that justifies the line item. If the answer is "projects that used to need an entire squad," the manager needs data to show it.

How do you track scale, productivity, and cost in a model where the unit is a senior person + AI instead of a squad? AI has a real cost, sometimes out of proportion to a simple task, and the first honest analysis will probably show that this format works for specific projects and not for the whole operation.

The risk of getting the metric wrong is concrete. If the ROI ends up weak, the HI-C becomes an expensive line item nobody defends internally, and at the next cost review it's the first one cut because its value isn't legible, regardless of merit. A manager who can't show what this person produces relative to cost, including the cost of inference, is building a role that will collapse at the first sign of trouble, even if the work is real, and it is.

There's also the other side, which worries me less but is still there: overestimating the model and generalizing too early. That means swapping a squad for a HI-C on a project that needed a squad, then finding out too late that the gain was limited to specific cases. Both mistakes are costly, and the second costs more because it means dismantling something that works.

All of this is an open door, and there's no ready-made model

What's happening is a door AI opened in how tech teams are organized, rather than a defined model with a ready-made playbook. Behind it are other doors that I'm sure we can't see yet. HI-C is one possible form, and others will emerge. The question "what's the right new org chart?" is probably badly framed, because it assumes there's only one and that it will hold still long enough to be studied.

What I like about this moment is that it's been a long time since tech management had a structural reorganization whose outcome was genuinely open. The last ten years were variations on the same theme: the agile squad, the tribe, the pod, the Spotify model, Team Topologies, all derived from the same premise that the group is the unit. Now that premise is being questioned for real, and that's rare.

I'm not ready to reorganize my teams around HI-C. I don't have the metrics, and I don't have the control I'd need. What I do have is the sense that treating this as a career trend is dodging the problem. The structural question is whether the company is organized to absorb these people or to waste them, and almost every company I observe still has ground to cover here, out of structural inertia rather than bad faith.

The unit of production changed before the unit of management. I honestly don't know how that gap will close.