A family is not a risk score.
Yet most plans still start with one: a questionnaire, a number, a model portfolio. Moondog lets you start with the family. It listens to each family member in private, learns what the money is really for, and builds a plan for that family and no other. The plan is anchored on the loss they can live with and tested against 2008, the pandemic and the 2022 rate shock. Every number calculated, never guessed. You sign. They understand. And every family gets your best judgement.
Built for the firms that serve significant wealth
Exceptional advice, held together by hand.
- The method that sets you apart lives in one person's head. It cannot be insured, sold or handed on.
- Scenarios built by hand in spreadsheets rarely reproduce, and rarely audit, which is exposure the day a regulator asks to see the reasoning.
- Decisions are made in meetings that leave no formal record.
- A polished report takes days, so the next family waits, or the work is turned away.
- Intentions are captured once and structured into a clear policy.
- Returns, risk and stress come from a single deterministic engine.
- Every family decision is recorded, versioned and approved.
- Word and Excel documents follow from the same engine figures.
From a family's intentions to a strategy they can stand behind.
Capture
A guided questionnaire turns conversations into a structured investment policy, with a review gate you control.
- ✓Objectives & horizon
- ✓Risk tolerance
- ✓Liquidity needs
- Constraints & exclusions
- Tax & jurisdiction
- Governance & reporting
Advise
You review the answers, settle where the family differs, and approve the profile and the policy.
How much of the portfolio should stay liquid within 12 months?
Model
A deterministic engine produces three comparable scenarios: Sleep Well, Core and Dynamic.
Agree
The family endorses one direction. The decision is recorded, versioned and ready for the report.
Every allocation traces back to policy.
| Equities | 0% |
| Fixed income | 0% |
| Private markets | 0% |
| Real estate | 0% |
| Cash | 0% |
Illustrative allocation. Sample figures, never a real client.
What the family actually receives.
Investment profile and scenarios
Family ref. UHNW‑0247 · v3 · approved
- Objective
- Preserve capital across three generations
- Horizon
- Perpetual, reviewed every 10 years
- Risk capacity
- Moderate · max drawdown −18%
- Liquidity floor
- 0% accessible within 12 months
Illustrative output. Sample figures, never a real client.
The model handles the language. A tested engine handles every number.
Built to stand up to scrutiny.
Hosting in Frankfurt, a strict boundary around the model, and a record of every meaningful action.
Hosted in Frankfurt
The platform and its database run in Frankfurt, and the default AI provider is European.
Pseudonymised by default
Family members' names are replaced with pseudonyms before every model call. The exact payload is retained as proof.
A tamper-evident record
Every meaningful action is logged and time-stamped, built so the record cannot be quietly changed.
The ground is shifting under hand-built advice.
Scrutiny is rising
MiFID II and BaFin increasingly expect to see the reasoning behind advice, not only the recommendation.
Clients expect to see the work
The next generation, raised on transparency, wants to understand how a number was reached before they trust it.
Wealth is changing hands
More families are passing on capital than at any time in living memory, and deciding who advises them next.
Why not just keep doing it by hand?
Excel and a junior analyst
Spreadsheets do not reproduce and do not audit, and the method leaves when the analyst does. You are renting capacity, not building an asset.
A general AI assistant
A general model will confidently invent a number. Moondog's figures come from a tested engine, and the model never touches the maths.
Build it in-house
Auditable, EU-hosted software is years of engineering. That work is already behind us.
Judgement, encoded by people who have built serious systems before.
A partnership, not a purchase.
Questions, answered.
Does the AI decide returns or risk?
No. The model only structures language. It reads prose, drafts a policy and proposes questions. Every figure comes from a deterministic engine, and a person reviews it before a client ever sees it.
Where is our data held?
In the European Union, in Frankfurt. The model sees only pseudonymised inputs, and the default AI provider is European. Every meaningful action is written to an append-only audit trail.
Does Moondog replace our methodology?
No. It embeds your methodology rather than replacing it with a generic algorithm. The model tailors at most a few questions, and you approve the policy before any scenario is built.
How does this support our regulatory obligations?
Reproducible figures and an append-only audit trail are designed to support a MiFID II suitability record.
Is this website the platform?
No. This is our public site, and it holds no client data. The platform is a separate, access-controlled application for advisers and their clients.
See Moondog on your own scenarios.
A considered, 30-minute introduction for advisory firms and family offices.