Let your clients’ AI check how you decide

Your AI turns past decisions and their reasons into a Decision Model. You review it and choose what to publish.

Found in your past work

  • Keep the name. Put the CMO's energy into the one-line category sentence that sits under it.

    Approved

  • Labels back under every icon. If five labels do not fit, there are too many tabs.

    Approved

  • Publish the prices of the two self-serve plans above the fold on the pricing page, and keep "contact sales" only for the enterprise tier.

    Stays private

  • The founder picks, today. Both work; the argument already costs more than either color.

    Approved

Your AI drafts the list. You approve what leaves and cross out what stays private.
  • Your AI checks with you five timesthen a short form to publish
  • Your raw sources never reach avvaonly the decisions you approved, redacted
  • You choose who can open itprivate, your company, or the marketplace

What you get

Start with a private model. Publish it when you trust it.

  • In your own AI

    Add your private model to the AI you work in. It reads how you decide before it answers you.

  • On a public page

    A sample of your decisions, with your reasons, for people who have not worked with you yet. The page can link to where they book you.

  • In your clients’ AI

    When you publish new decisions, their AI gets your updated model on its next request.

From your work to their AI, in a 45-second film

You approve the decisions. Their AI can consult them on new questions.

Noor gets the same questions every week. Her own AI finds the decisions she already made, she approves each one, and the model stays private until she chooses otherwise. Then someone else’s AI reads it before it answers. No sound needed. Read the film as text.

What their AI reads

A system prompt can list your rules. Past decisions show what you checked first and where you stopped.

  • How you approach a decision, what you weigh first, and where you stop
  • Your decisions, in English, with your own thresholds kept and deal figures removed
  • Your reason for each where you gave one, what else was in play, and what you ask first

An example case

  1. What decided itBuyers already know what this category looks like · The difference has to be said, not drawn

    Series A fintech, 20 people, 300 paying businesses. The board said the marketing site "looks dated" next to two competitors and asked for a rebrand — new logo, new palette, a custom illustration system — before a Series B roadshow in ten weeks.

    No rebrand and no new logo. Rewrite the homepage headline, replace every illustration with real product screenshots, and move the site's type and color closer to what the category leaders use, not further away.

    Why

    In finance a buyer reads "looks like a bank" as "safe", and that trust is free if you borrow it. The competitors look newer because they show their product and say one sharp sentence; their marks have nothing to do with it. Difference belongs in the sentence.

One public page and one link. Your client adds the link to the AI app they already use, and their AI reads how you decide before it answers them.

How it gets built

From the AI you already use, the one that already sees your work.

01

Add avva Studio to your AI

One link. Your AI finds decisions you made, and your reasons, in what it already sees.

02

Approve the list before anything leaves

It saves what it found on your side. Nothing reaches avva until you say so.

03

Your AI builds the model

It works out what each decision turned on and where you stop, and checks the draft against them.

04

Hear it back, then publish

It reads the model back and takes your corrections. Private by default; sign in only to publish.

What stays private

avva never receives your raw source material.

  1. Your AI reads your sources

    Inside your own Claude or ChatGPT account.

  2. You cross out what is private first

    Saved on your side before any text moves. Cut anything personal, client-owned or under NDA.

  3. Your AI takes identifying details out

    Names, people, ids and deal-identifying amounts come out. Your own thresholds and the shape of the situation stay.

  4. Only what you approved reaches us

    We store the finished model, including your cases (the situation, what you decided, your reason) as you approved them, and keep no copy of the approved file.

Who can open it, and what they see

Checked on every request. Update, unpublish or delete it any time.

Private

Private to your account. Our operator can read it through a support tool that logs each read.

Your company

Invited colleagues only. To anyone outside, it does not exist.

Marketplace

A public page and a link anyone can connect. Or unlisted: only the link finds it.

  • On the page: who you are, what you cover, five decisions you pick. Never the sources themselves.
  • In their AI: your published model. Their AI writes the answer. It is not a live consultation or professional advice.
  • Who runs it: its founder. There is no company yet, so nobody can sign a data-processing agreement on avva’s behalf. Details
  • Usage: requests, separate sources and active days, in your dashboard and nowhere else.

Put it on your site

Visitors try your Decision Model in their own AI before they hire you.

The card

One line from your dashboard: your photo, what you cover, and a button into your page.

Free during beta

Use your own AI. No account until you publish.