Skip to content
Brain

Teach it once.
Every ad remembers.

Most teams explain their brand over and over: to a freelancer, to a new hire, to whichever AI tool they opened this morning. The Brain is where that explanation lives instead. You teach it once, and it reaches every agent, every teammate, and the assistants you use outside Orisu.

Website in · brand book in · imported memory in
3ways in: your website, your documents, an assistant's memory
1brand per workspace, on purpose
Everymemory cites its source and who signed off
0credit cards needed to start
§ 01 · Fill it

Three ways in. You never fill in a form.

Point it at your site, drop in a document, or paste what your assistant already knows. Every fact it takes in remembers where it came from.

  • Website. Palette, fonts, voice and claims, plus your product photography, kept as reference.
  • Documents. A brand book, a past campaign, a spreadsheet, read page by page.
  • Imported memory. One prompt for ChatGPT or Claude; paste the reply back and Orisu reads the brand out of it.
§ 02 · Review

Nothing it worked out on its own goes live unread.

Anything Orisu inferred arrives unconfirmed and waits in Needs review, with the facts it is most sure of first. Confirm what is right, dismiss what is wrong, and edit what is nearly right. A correction counts as a confirmation, because a person wrote the words.

Every memory shows who taught it and who signed off. Dismissed facts stay in history, so you can see what the Brain used to believe and when that changed.

§ 03 · Words

The most checkable rule a brand has.

Mark a word Never and your agents will not write it. Mark one Preferred and they reach for it instead. Careful is for a word that needs a specific claim behind it.

Your guidelines probably say all this in prose already. A model follows “never write cheap” far more reliably than a paragraph that mentions it in passing.

§ 04 · Use it

Everywhere you make things.

Teach it once and it is there wherever you work.

  • Copilot. Starts every message knowing your identity and word rules, so launch copy comes back in your voice.
  • Canvas. The Brain node hands your rules and references to any step.
  • Feed. Every row carries the same brand.
  • MCP. Your own Claude or ChatGPT can read it, search it and add to it. Its proposals land in your review queue.

Explain your brand for the last time.

Paste your URL and the Brain does the first pass. You do the review. Free to start, and the free tier needs no card.

Good to know

Before you teach it anything.

Will it overwrite what I already wrote in the profile?

No. A profile field you wrote by hand is never overwritten by an import, and the words you banned are kept and merged rather than replaced.

What if it gets my brand wrong?

Anything it inferred waits in Needs review until a person confirms it. Edit the wording or move a fact to a better category; saving a correction confirms it at the same time.

I run several brands. Where do they live?

One brand per workspace. Give each brand its own workspace so an agent generating for one is never handed another's rules. Encoding a second website into a workspace that already has a brand asks you first.

Can I turn it off?

Yes. The Brain switch beside the copilot composer turns it off for a conversation where you would rather work without it.

Teach it once.

Paste your URL, review what it learned, and run your first on-brand agent today, free, no credit card.

No credit card · Your brand stays in your workspace · Cancel anytime