For organisations
Put the instrument to work.
Score your people's instinct. Query the trained Metamodel on your own questions. Or study the phenomenon with us.
Every sealed call teaches the model something.
When human instinct holds, when it doesn't, and under which conditions.
Three ways to use what it learns. Score. Query. Study.
One instrument. Three ways in.
Start with your people, your questions or your research.
01 The Audit
Score your people.
Capture your team's instinct before the reasoning, then score it against outcomes and against chance.
Founding clients02 Metamodel access
Query the model.
Put your organisation's questions to the trained Metamodel. Get a probability, its range and its track record.
Early access by application03 Research access
Study the phenomenon.
For universities and labs. Use the Metamodel to study precognition, then design experiments we build into the platform.
First partners from 202701 The Intuitive Intelligence Audit
Score your people.
Forecasting tools score the considered view. We capture the instinct first, separately, then score both.
Five weeks.
Seventy-five calls each.
Three short calls each working day, per person. Enough for individual scores to mean something.
Week 0 · Assess
We agree the question, the domains and the consent model.
Weeks 1 to 5 · Collect
Your people make quick calls before the analysis. Then their considered view.
Week 6 · Analyse
Both are scored against outcomes, against chance and against each other.
Week 7 · Deliver
You get the readouts, and what to change in how decisions get made.
What the Audit gives you
Five readouts.
Sold in order of evidence. Each one carries its call count, so a claim on eight calls is held more lightly than one on eighty.
- Instinct deltaDoes your team's first instinct beat its considered view, or the reverse?
- Calibration scorePer person and domain, shrunk so luck can't top the table.
- Edge over consensusWhere your people diverge from the market, and whether it pays.
- Conditions mapThe times, states and speeds where each person's calls hold up.
- Signal indexGated. Residual accuracy after controls, past thresholds set in advance.
02 Metamodel access
Query the model.
Put your organisation's questions to the trained Metamodel. It answers with a probability, its uncertainty, and how far that sits from chance.
Ask a question that will resolve.
A market shift. A launch outcome. A position your organisation needs to take.
Many calibrated human calls. Weighted by the conditions that predict accuracy.
More sessions. Narrower answer.
Uncertainty shrinks as resolved calls build up. Illustrative.
What you get
An answer you can weigh.
Use it as one input when you take a position. Never as the only one.
- A probability and its range.Not a verdict. The width tells you how much weight it can bear.
- Distance from chance.Every answer shows where it sits against a coin flip.
- A full track record.Every past answer is resolved, scored and shared with you. Edge is claimed only where it has been shown.
- Question design.We help frame your questions so they resolve cleanly and can be scored.
Early access by application, opening as session volume allows. Outputs are statistical estimates, not financial or investment advice. See the Disclaimer.
Where instinct already decides
Wherever the call comes before the data.
Score the people who make those calls. Or ask the model for a second, calibrated view.
Investment committees
Does the room's first read beat the memo? On which kinds of deal?
Strategy and foresight
Who sees a shift early, and how calibrated are they when they do?
Risk and crisis
Fast calls under pressure, scored against what actually happened.
Senior hiring
First impressions against eventual performance, measured not assumed.
03 Research access
Study the phenomenon.
For universities, institutes and independent labs. Use the Metamodel to study precognitive phenomena at scale, then propose experiments we build into the platform.
- Query the model.Explore population patterns across consenting participants, in groups of at least thirty sessions.
- Test a hypothesis.Work with anonymised exports under a data sharing agreement. No names, no free text, no exact times.
- Design an experiment.Propose a new protocol, game or trial type. We review it and build it into the platform.
- Run it properly.Pre-registered, sealed and blinded, with consenting participants.
- Publish.Jointly or independently. Null results included.
Free research licences for our first two academic partners.
Governance
Data handled like it matters.
- Team-level by default.Audit reports aggregate to the team unless each person agrees to named results.
- Consent first.Audit data is performance data about employees under UK GDPR. Everyone opts in, and can opt out.
- No body data.Audits never collect heart rate or any other biometric data from your people.
- Model answers carry their evidence.Probability, range and track record always travel together. Patterns stay exploratory until resolved questions say otherwise.
- Researchers get anonymised data only.Under a data sharing agreement that bans re-identification.
- No overclaiming.Signal index stays gated. If an edge isn't there, your report says so.
Pricing
Founding clients first.
Founding Audit
For our first three clients, in exchange for a case study.
Standard Audit
After the founding round. Scoped to team size and domains.
Metamodel access
Priced by question volume and term. Early access opens with session volume.
Research access
Licences for our first two academic partners. Agreed per study after that.
Every Audit seat includes Pro membership.
Start here
Talk to us.
Thirty minutes. Tell us what you'd like to test, ask or study, and we'll tell you honestly whether we can.
We'll reply within two working days.