Twenty years ago Ron and I split a small tech org in two. It was the first time I put on a Director title. I took Technology Development, he took Engineering. I worked in the chaos of ideas, and he turned the good ones into something we could ship.

Twenty years later, that divide and a few other familiar ones are what Boris Cherny named.

On June 28, 2026, Cherny (Anthropic, creator of Claude Code) posted what he sees on his own team now that engineering, product, design, and data science are melting into one kind of work. Five archetypes, not five job families:

  1. Prototyper: comes up with brand new ideas; churns out many, most of which don't ship

  2. Builder: quickly turns a prototype or idea into production-grade product or infra

  3. Sweeper: cleans up the UI, simplifies the code and system, unships, optimizes performance

  4. Grower: takes something built and iterates it toward Product-Market Fit

  5. Maintainer: owns a mature system so it stays secure, reliable, fast, and efficient as it scales

I was the Prototyper. Ron was the Builder.

I do all five in a given month. Those two are where I am actually awake: the late nights sitting with an idea and building a prototype just to show it, because agents make that cheap now, and the days you take one of those ideas and try to make it real enough that a customer can touch it.

The flood

When coding agents got good enough to use every day, prototypes stopped being a scarce art. They became a weather pattern.

They land faster than we can absorb them, and they are slowly making their way into production and touching real customers. At home I am still averaging around a billion tokens a month on vibe projects. How many vibe-coded apps have you built?

The first "it runs" demo arrives and turns heads. Getting that demo to where it can hold a customer is a different job with a different set of skills. Operating it after that is a third one.

I have spent most of my recent energy in Prototyper and Builder, and it felt like the skill that mattered, because the bottleneck was "can we even try this?" The bottleneck moved. Now we are hitting the shift from build to support, and the skills do not transfer. Now the job is to operationalize.

It has always felt like a different mode of operation. Now it feels like a different job. Operations was never optional for engineers, and embedding DevOps in engineering teams is an old pattern, one I learned in my Amazon days. What is new is how much of it one person can hold. Could a single engineer now operationalize something that used to take a full team?

Cherny's own post says the mix is the point, not picking a favorite archetype and keeping it:

❝

A healthy team needs a mix of these, depending on the product:

  • A product that is new and pre-PMF needs people that are strong at 1+2+3

  • A product that is growing and has found PMF needs 2+3+4 and some 5

  • A product that has strong PMF needs 3+4+5 and some 2.

Maybe product roles of the future will look more like this, and less like the domain-specific roles of today?

Boris Cherny, on Threads and X, June 28, 2026

That is a staffing map, not a personality quiz.

Titles were already melting

A month earlier, on Platformer with Casey Newton, Cherny was already talking past "software engineer" as the durable label:

❝

I don't think we're going to call them engineers… But if we talk about people writing code, or using agents to write code, I think there will be 100 times more engineers than there are today.

Boris Cherny, Platformer interview with Casey Newton

My bet is on the return of the Member of Technical Staff. At Riverbed Technology, my first job in Silicon Valley, everyone in engineering carried the same official title: Member of Technical Staff. We had product managers, but they sat apart from us and stayed high-level and strategic.

No extra layers of non-technical support staff. The engineering leaders built the roadmaps, and they built the prototypes of new products that then changed the roadmap. I was too junior to know where the ideas originated, but I rubbed elbows with some of the best technical teams I have ever worked alongside. Teams of engineers built those prototypes and new concepts, and then engineering refined them into production. Shout out to those Riverbed colleagues, still one of the best groups of future VPs, founders, and leaders I have had the pleasure of working beside.

We have all heard the forecast by now, usually truncated to whichever half suits the person quoting it:

❝

One is that a lot of companies will need fewer engineers, because each engineer is more productive… At the same time, a lot of companies will need many more engineers, because every engineer is more productive — the company can do more things, start more products…

Boris Cherny, Platformer interview with Casey Newton

Fewer and more, in different companies, at the same time. The title melts. The modes stay.

Then there is the line that gets quoted at four words, with the half that matters cut off:

❝

Coding is solved for the kinds of coding that I do… When you think about what engineers do, coding is a small percentage of it… when the model does the coding, they're freed up to do all the other stuff they actually enjoy, like talking to users and figuring out what's next.

Boris Cherny, Platformer interview with Casey Newton

That "other stuff" is where Sweeper, Grower, and Maintainer live. Unshipping. Grinding toward PMF. Making a mature system boring. The prototype flood does not delete that work. It fills the queue with unfinished versions of it.

That part I want. I love talking with users and figuring out a solution to a problem in their lives, giving them something they did not know to ask for. And I am sure some of you read that and think it sounds like a horrible role, but are excited by the other roles Boris sketched out.

What am I getting wrong?

Builder mode is agents writing, PRs stacking, a demo that makes somebody lean toward the screen. Support mode is the alert, the customer who found the ugly edge, the prototype that shipped further than anyone meant it to.

In the last month I have named this shift out loud to one of my teams, nine months into building a new platform. We consciously switched from Build to Maintain and Grow.

According to Cherny, the same person can span two or three of the five. He added that the archetypes are not tied to job function: on his team some designers land in slot 1, some in 2, some in 3, and the same holds for engineers and PMs. The mistake is anchoring on yesterday's title while the product stage has already moved underneath you.

If your week is still mostly Prototyper and Builder and your product is already in customers' hands, you are late to the skill shift, not early to the idea shift.

The rule

When prototypes are cheap, scarcity moves downstream: production hardening, unshipping, PMF iteration, keeping mature systems boring. Name the mode your product stage actually needs, then check whether it is the mode you are any good at.

Where are you going to thrive? Which role gets you into flow?

References

The five archetypes are Boris Cherny's, posted June 28, 2026: Threads · X. The reel that put the list in my feed is @itsmariahbrunner, July 3, five days later. The list is his. The amplification is theirs. If you share the clip, credit the June 28 post.

  • Platformer on jobs and "builder": article · YouTube

  • Pragmatic Engineer on prototypes over PRDs: write-up · YouTube

  • Business Insider summary of the June 28 post: BI