Skip to content
AI-Native PM
Field Notes

Field Note

How We Used Agent Fleets to Build This Site

The picture of building with AI is one person and one chat window. This site was built the other way, with many workers running at once and a person setting the brief and judging what comes back. That is the real shape of the work once the scale is real.

· 5 min read

The common picture of building with AI is one person and one chat window, asking for a thing, reading the reply, asking for the next thing. It is a tidy picture, and it is not how this site got built. Most of the heavy work here ran as many workers at once, each handed its own task, with the two of us setting the brief and judging what came back. The chat window was the smallest part of it.

Take a few of the jobs behind what you are reading. Every chapter in the library cross-links to others by their titles, and when the structure changed, those links had to change across the whole set at once. A batch of older diagrams, carried over from an earlier site, had to be redrawn in the current house style. A whole part of the course had to be drafted, chapter by chapter, against the same voice. When dark mode went in, every diagram on the site had to be moved off fixed colors and onto the theme tokens so it would read in both palettes. None of these was typed out in sequence in a single thread. Each one went to a group of workers running in parallel, one per file or one per chapter, finishing at the same time.

One brief, many parallel workers, one verified resultA single brief on the left fans out to four workers running at once, each on a different task: draft a chapter, build a diagram, re-link the text, audit the voice. Their work converges on the right into one step, verify then keep, where a human judges what to keep. Caption: one brief fans out to many workers, then one judgment keeps what is good.HOW THIS SITE GOT BUILTSET THE TASKone clear briefDraft a chapterBuild a diagramRe-link the textAudit the voiceALL AT ONCEVERIFY, KEEPone human judgesOne brief fans out to many workers, then one judgment keeps what is good.

The moment one worker becomes many, the work changes form. The bottleneck stops being how fast you can produce, because production is now happening in parallel and cheaply. It moves onto two harder things. The first is stating the task clearly enough that a worker who cannot stop to ask a follow-up can still finish it correctly. The second is judging what comes back, because a group of workers will return a group of plausible answers, and plausible is not the same as right.

Building with AI at any real size is orchestration, not conversation. You stop typing the work and start briefing the workers and judging what they return.

That is the whole job once the scale is real. A vague brief produces a fleet of confidently wrong files, all at once, which is worse than one wrong file at a time. A sharp brief, the kind that names the constraint, the format, and the test for done, produces work you can actually keep. And nothing a fleet returns ships on its own word; every batch met a second pass that checked it, which is its own note, How We Caught AI Mistakes With a Second Agent.

This is not a trick of building a website. It is the same shift the Frontier of The Builder's Stack teaches as orchestration: running many capable workers in parallel and keeping a person at the point where the judgment happens. The skill that mattered most across this build was not writing a better request. It was decomposing a large job into pieces a fleet could take in parallel, and then reading the results well enough to keep the good ones and send the rest back.

Where this goes next

The discipline behind this is the subject of the Orchestration part in the Frontier of The Builder's Stack, which covers how to split a job into independent pieces, run them at once, and verify the results before you trust them. For where this sits in the day-to-day loop of running an AI product, read The Operating Manual. The companion habit that keeps a fleet honest, a second worker whose task is to refute the first, is in How We Caught AI Mistakes With a Second Agent.

Sources