Cerevia live — walking the city, the council chamber deliberating, and the studio. Not loading? Some networks and VPNs block YouTube — open it directly.
Brand in. Calendar out. Numbers back in.
Four stages, and the last one feeds the first. Open any stage.
Point it at your own site. It reads the live pages — name, tagline, colour, logo, the social accounts you actually link to, and what your headings say you do — instead of a questionnaire you fill in once and it forgets. Nothing is inferred or guessed: if a signal is not findable it is reported missing rather than filled in with something plausible. Forty-one charters across seven industry profiles mean it arrives already knowing what matters in your category.
A month, not a post. The calendar is built against your brand and the goal you actually have — lead generation and thought leadership are not judged by the same numbers, so they are not planned the same way. You stop opening an empty month.
Drafts are written against that brand and that plan, with two or three hook variants, each tagged with its attributes: hook type, format, call to action, pillar, length, posting hour. Then a council reads it — sized to what is at stake.
You report what the numbers did. The system pairs the outcome to the decision that produced it and updates what it believes — or refuses to, and tells you why. This is the stage every other tool is missing.
Every draft goes through a governed decision.
The councils are not there to slow you down. They are the quality bar that stops the loop drifting into generic AI output over months of iteration — and the reason the system can tell you why something worked, not only that it did.
Deliberation is routed by what is at stake. A caption convenes brand and context; a campaign launch convenes far more. A single post never convenes the whole house — that would be slow, and it would be theatre. Where seats disagree, you see the disagreement rather than a smoothed-over average.
Context adapts expression; the reasoning stays constitutionally invariant. In plain terms: it should sound different on LinkedIn than on Instagram, and different again in a regulated category — while the standard it is held to does not move an inch.
Underneath every one of those sits the same ten-stage loop.
Something happens. A customer says a thing, a campaign lands, a competitor moves. The work starts before anyone has decided what it means.
The signal is recorded as evidence — what was observed, when, and by whom — kept separable from anyone’s reading of it.
Evidence accumulates into claims the system is willing to stand behind, each one still traceable back to what it came from.
Knowledge is held so it can be found again. Not a document graveyard: the reasoning stays attached to the conclusion it produced.
Competing readings are argued out in the open. A red team asks what would have to be true for this to be wrong.
A choice is made and recorded with its dissent — what was decided, on what evidence, and who disagreed at the time.
The decision leaves the system and meets the world, carrying its record with it.
What actually happened is written down against what was predicted. Recording the prediction first is mandatory — a system that never wrote down its guess cannot discover it was wrong.
The gap between the two is the lesson. It attaches to the reasoning that produced the decision, not to the person who made it.
Rules that keep failing get changed. The constitution itself is revisable — in the open, with a record of why.
What it actually looks like inside.
Four views of the same system: the constitutional connectome that governs every decision, the evolution and execution networks, the learning network that turns outcomes into capability, and the impact centre.
These are the real working diagrams, not marketing illustrations. Open one and follow a path through it.
Fair warning: they are drawn at organization scale, and they use organization language — institutional memory, human ROI, council activity. That is deliberate. The runtime governs decisions generally; social is where we start, because that is where the volume of decisions is and where the feedback comes back fastest.
A decision that survives the month it was made in.
Every deliberation leaves a record: what was decided, on what evidence, who dissented, what was predicted, and what actually happened. Not a log line — a document you can hold up to the light six months later, when you are trying to work out why the thing that worked worked.
Drag to turn it.
Build the next chapter with us
Cerevia is early. The founding circle shapes what it becomes.