How to read this catalogue
This is a private, internal strategy tool — a structured way to explore, compare and shortlist more than one hundred AI business opportunity ideas. It is not a public marketing site and it makes no promises.
Ideas are hypotheses
Every idea here is a hypothesis, not a proven business. Nothing on these pages has been validated with paying customers. Read each page as "here is a bet we could test", never as "here is a business that works".
To keep this honest, the catalogue distinguishes:
- Facts — statements that could be checked against evidence.
- Market information — competition, pricing and adoption; changes over time, verify before relying on it.
- Strategic judgement — reasoned interpretation, clearly a point of view.
- Assumptions — things believed but not yet validated.
- Speculation — possible futures that cannot be known now.
Where a value has not been assessed, the interface shows Not assessed rather than inventing a
number. Pricing is always a hypothesis. Market scores are judgement. Legal and regulatory
notes require professional verification.
What the labels mean
| Label | Meaning |
|---|---|
| Status | Where an idea sits in its lifecycle. Everything starts at Concept. |
| Validation | Evidence collected so far. Everything starts at Not tested. |
| Market chips | Geographies the idea might suit (a starting guess, not a claim of demand). |
| Demo available | A frontend-only clickable prototype with synthetic data exists. |
| Not assessed | The analytical field has deliberately not been scored. Score it yourself from real evidence. |
Data and safety notes
- Every demo uses synthetic data only and is clearly labelled "synthetic demonstration, not a production system". Demos avoid medical, legal, structural, safety and financial decisions.
- Healthcare ideas are limited to administration — never clinical decisions.
- Municipal ideas stay away from automated rights, benefit, permit, enforcement or eligibility decisions; a human civil servant always decides.
- Construction calculations and compatibility rules stay under qualified human and deterministic control.
How to use it
- Skim the homepage for the thesis and the top candidate ideas.
- Filter the Catalogue by category, market or "has a demo".
- Open an idea, read its consistent 18-section page, and try any linked demo.
- Shortlist promising ideas (saved in your browser) and compare up to four.
- Pick one to test and open the Validation workspace.
- When a few ideas earn real evidence, assemble a publish set for a future public site.
:::tip The only validation that counts A customer's money or a credible procurement step. Interest and praise are not validation. :::