If AI has disappointed you, ask one question: did you brief it like a hire, or visit it like a website? This kit is the brief. One hire, the Profile and the Rules, one first job.
Free. No code. No patient data, ever. We email you a link, you create a free account, and the walkthrough opens.
A Co-CEO is not a website you visit. It is one workspace, a Claude Project or a ChatGPT Project, that carries your context and works three jobs for the practice owner. You hand the jobs over one at a time.
The recurring letters, emails and posts, in your voice rather than a chatbot's. This is the first job, because the output is easiest to judge.
A second opinion that knows your practice. It argues the other side of a decision before you commit, instead of agreeing with you.
Recommendations with reasons, under the Three Laws below. It never carries a decision alone, and it never assumes a yes.
There is no code in any of this. And no patient data goes into it, ever. Identity and voice, not records.
Run the Research Prompt in every AI research tool you can reach for free: ChatGPT, Gemini and Perplexity. Confirm or strike each report, then merge what survives into one file, practice-research.md, the facts half of your Profile.
Run the Voice Prompt on three real things you wrote and sent. Fill the three boxes, one sample each, and save what comes back as voice-notes.md, the voice half of your Profile.
Attach practice-research.md and voice-notes.md (from steps 1 and 2), run the Profile Builder, and save KB-01-Practice-Profile.md, knowledge base file number one. Step 5 adds it to the project.
Give it a seat in Claude, ChatGPT or Gemini (a Project on Claude or ChatGPT, a Gem on Gemini). Name it Practice Co-CEO and leave it empty. Step 5 puts the hire in it.
Paste the project instructions, then upload your Profile from step 3 and the Operating Rules, downloaded as Markdown from the accordion.
Start a new conversation and ask it the test question. The walkthrough has the exact wording and what each answer means.
The hire is ready. Give it one piece of writing you already redo, briefed in four lines, and judge the draft.
Give it the plan if you already have one written down, and bring it a decision if you have one sitting. Neither is a numbered job.
This map is free to read here. The prompts, the test wording, and every file to copy live in the walkthrough.
Three tools the walkthrough installs with the hire: which jobs to hand over, how to brief any job, and the rules that keep deciding safe.
Three questions. Run every candidate task through them before you delegate it.
If the task happens again and again in roughly the same shape, it is a candidate.
If you can write down what a quality output looks like, you can brief for it.
If the judgement can be written down and handed over, your Co-CEO can hold it.
Yes to all three: hand it over. Anything less: keep doing it yourself, or hire a human.
The hire is done: Profile built, Rules uploaded, test answered with your practice's own name. Now pick the job. The recurring letter, email or post you rewrite every time.
Feed it real work. When a draft misses your voice, fix the Practice Profile (KB-01), not just the draft. The document is the memory.
Bring it a real decision you are sitting on. Ask it to argue against your position before you commit either way.
Hand it one small, low-stakes decision to recommend on, with the Three Laws in force.
Read the week's work cold. Keep what earned its place. The next job waits for day 30.
One Co-CEO. One first job. Thirty days. Judge by the output.
This afternoon is the hire: a working Co-CEO and one first job. When the Profile is no longer enough, Your AI Think Team builds the full identity and voice with you, then the specialists, one job at a time. None of that is this month's work.
The walkthrough is free. We email you a link, you create an account, and the hire begins.
Later, if you want the guided six-week install: Your AI Think Team.