Industries
Four sectors,
chosen by consequence.
Not by budget.
We aim at the systems humanity and the planet cannot afford to lose. That is a selection rule, not a slogan: we choose by what breaks if it goes wrong and what compounds if it goes right. Attractive markets and essential ones are not the same thing.
The selection rule
One industry at a time, learned properly, before the next one.
We take one sector, learn it until we can argue with a practitioner, build what that sector actually needs, and only take it to market once it stands up. Regulated finance and legal came first and produced Sentinel and Citadel. Food and local resilience is where we are now, and produced Backgate. Carbon and waste is on the Lab bench today.
We are deliberately not everywhere at once. Energy and land are where the same logic points next — that is a statement about our reasoning, not a claim that we serve them.
Where we say no
Healthcare is analysis only. The questions are live for the people working in it, so we publish what we are working through — and we take no clients there yet. We would rather say that than imply a practice we do not have.
The four
Where it fails, and what that costs.
Each page sets out the sector’s consequence line, what we have actually built there, and what we will not claim about it.
Where capital stops if it fails
Legal & Financial Systems
The FCA, the PRA and the SRA have not written new AI rules — they are applying the ones you already answer to. The sector Sentinel and Citadel were born in: the complete inventory, the exposure priced in pounds, the evidence kept live.
Where it fails in days, not years
Food & Local Resilience
The same machinery that governs AI inside a bank, built for a shop with three staff. Food leads because food is the business a place cannot do without — and the business least able to buy its way into modern technology.
Where the damage compounds
Carbon & Waste Management
The hard part of a sustainability claim is not making it — it is showing, later and to someone sceptical, exactly what it rests on. Carbon Ledger asks how a claim survives challenge years later. We certify nothing.
Where the risk concentrates
Healthcare
AI is already in clinical workflows, and the governance questions are live for the people working in it. We publish the analysis for CQC-regulated providers. We take no clients here yet, and we say so plainly.
What the method does not change
The sector decides what counts as proof. The discipline does not move.
A regulator wants an audit trail; a farm shop wants the till to balance. What counts as evidence is different in every sector we enter — but the named owner, the evidence state on every claim, and the working system at the end are the same on all four pages.
Not one of the four? Tell us anyway.
Thirty minutes with a founder. If your sector is not one we have learned properly, we will say so — and tell you whether the method still transfers or whether you want a specialist.
Price agreed before we start · Founder-led · NDA available