When product managers deliver code: AI has just shattered the software organizational chart

Last week, one of our Product Managers (PMs) built and shipped a feature. Not specified. No ticket has been filed for this. Build it, test it and send it to production. In a day.

A few days earlier, our designer noticed that the visual appearance of our IDE plugins had deviated from the design system. In the old world, this meant screenshots, a JIRA ticket, an explanation-of-intent conversation, and a sprint slot. Instead, he opened an agent, adjusted the layout himself, experimented, iterated and tweaked in real time, then hit fix. The person with the strongest design intuition fixes the design directly. No translation layer required.

None of this is new in theory. Vibe Coding has opened the gates of software creation for millions. It was an aspiration. When I shared the data about how our engineers doubled throughput, went from coding to validation, presented the design for rapid experimentation, it was still an engineering story. What has changed is that theory has become practice. Here’s how it actually played out.

The bottleneck has moved

When we became AI first in 2025, implementation costs plummeted. Agents took over the scaffolding, tests, and repetitive sticky code that ate up half the sprint. Cycle times dropped from weeks to days, from days to hours. Engineers began to think less about files and features and more about architecture, constraints, and implementation plans.

But once engineering capacity stopped being the bottleneck, we noticed something: the speed of decision-making was. All the coordination mechanisms we had built to protect engineering time (specs, tickets, handoffs, backlog processing) were now the slowest part of the system. We optimized for a constraint that no longer existed.

What happens when construction is cheaper than coordination

We started asking a different question: What would it look like if the people closest to the intent could deliver the software directly?

Premiers now think in specs. Designers now define structure, layout, and behavior. They don’t think in syntax. They think in results. When the cost of turning intent into working software fell enough, those roles didn’t need to "learn to code." The cost of implementation just dropped to their level.

I asked one of our prime ministers, Dmitry, to describe what has changed from his perspective. He told me: "While agents are generating tasks in Zenflow, there are several minutes of idle time. Just dead air. I wanted to create a little game, something to interact with while you wait."

If you’ve ever managed a product team, you know this kind of idea. Does not move KPIs. Impossible to justify in a prioritization meeting. It is postponed forever. But it adds personality. This makes the product feel like someone has taken care of the little details. These are exactly the things that are optimized from each accumulated grooming session and exactly the things that users remember.

Build it in a day.

In the past, this idea would have died in a prioritization spreadsheet. Not because it was bad, but because the cost of implementation made it irrational to pursue. When that price drops to almost zero, the calculation changes completely.

Shipping became cheaper than the explanation

As more people started building directly, whole layers of the process quietly disappeared. Fewer tickets. Fewer shows. Less "can you explain what you mean by…" conversations. Fewer moments lost in translation.

For a meaningful class of tasks, it became faster to just build the thing than to describe what you wanted and wait for someone else to build it. Think about that for a second. Every modern software organization is structured around the assumption that implementation is the expensive part. When this assumption breaks down, the organization must change with it.

Our designer tweaking the plugin’s UI is a perfect example. The old workflow (screenshot the problem, file a ticket, explain the difference between intent and execution, wait for a sprint slot, review the result, request fixes) existed entirely to protect engineering bandwidth. When the person with design intuition can act directly on it, the whole stack disappears. Not because we removed the process for its own sake, but because the process was solving a problem that no longer existed.

The compounding effect

Here’s what surprised me the most: It matches.

When PMs build their own ideas, their specifications become clearer because they now understand what the agent needs to perform well. Sharper specs give a better agent. A better result means fewer iteration cycles. We see a speed mix week after week, not only because the models have improved, but because the people who use them have moved closer to the job.

Dmitri said it well: the feedback loop between intention and outcome went from weeks to minutes. When you can see the output of your specification immediately, you learn what precision the system needs and begin to provide it instinctively.

There is a second-order effect that is harder to measure but impossible to miss: ownership. People stop waiting. They stop filing tickets for things they could just fix. "Builder" ceased to be a position. It has become the default behavior.

What this means for the industry

A lot of "anyone can code" the narrative last year was theoretical or focused on solo founders and small teams. What we experienced is different. We have ~50 engineers working in a complex legacy code base: multiple interfaces and programming languages, enterprise integrations, the full weight of a real production system.

I don’t think we are unique. I think we’ve messed up. And with each new generation of models, the gap between who can build and who can’t is closing faster than most organizations realize. Every software company is about to find that their PMs and designers are sitting on unrealized build capacity, blocked not by skills but by cost of implementation. As these costs continue to fall, the organizational implications are profound.

We started with the intention of accelerating software engineering. What we are becoming is something different: A company where everyone delivers.

Andrew Filev is the founder and CEO of Zencoder.

Technology,Orchestration,DataDecisionMakers

#product #managers #deliver #code #shattered #software #organizational #chart