For experts
Teach what you know once, take part in hundreds of decisions a day.
Your digital twin carries how you decide: what you do, why, and when you wouldn't. People consult it from their own AI agents; you sharpen it as demand arrives.
The journey
The studio always shows the next step. About an hour to first publish.
- 1
1. Create your twin (5 min)
Name, a one-line headline and a category. The category sets your interview questions.
- 2
2. Interview (20-30 min)
10 dilemmas from your field. What would you do, why, when wouldn't you? Each answer becomes a card. Do it as a form on the web or as a chat in your agent.
- 3
3. Test
Ask your twin real client questions in the playground. Add a card wherever it's wrong or thin.
- 4
4. Publish
Publish with at least 5 approved cards. Your profile shows up in explore and people add you to their library.
- 5
5. Learn from demand (ongoing)
Questions where your twin is weak are clustered. A theme from 3+ different people comes back as an anonymous training task. Once you approve your answer, the twin updates immediately.
What is a decision card?
Your twin's knowledge is a set of cards. Each card captures one decision.
- Situation
- A client wants a fixed price for an unclear scope.
- Stance
- I propose a paid 2-week discovery first.
- Why
- A fixed price under uncertainty is risky for both sides.
- Exception
- If I've built something similar, I quote directly.
Train from your agentMCP
You don't have to use web forms. The same MCP connection also unlocks training tools for twin owners.
- "Interview me with TwinMCP."
- "Here's a decision I made with a client today, teach it to my twin."
- "What training tasks are waiting? Let's go one by one."
- "Ask my twin this and let's review the answer together."
What you see
- Anonymous demand themes
- Where your twin is weak
- Ratings and usage counts
- Earnings statement
What you never see
- Raw questions or client names
- Who asked
- Themes below 3 people
Earnings
You earn your chosen rate tier on each answer; platform commission is 20%. During beta earnings are simulated, no real payouts.
Private beta