DataFab / Artificial Proprietary Intelligence
Artificial Proprietary Intelligence
The thesis, argued in full.
The first is an operating cost every competitor also pays, converging in capability and price, reaching your rival the same quarter it reaches you. The second already exists inside your organisation, available to exactly one buyer on earth, and impossible to license at any price. This is the route from owning it to using it.
The race everyone is watching
General intelligence is converging.
Enormous capital, extraordinary people, real progress — and no argument from us. But general means available to everyone, on the same terms, in the same week. Whatever is built will reach your competitor when it reaches you.
General intelligence
Learned from the world’s text. Identical for everyone who licenses it. Improves on a vendor’s schedule, not yours. A line of cost that buys parity, never advantage.
Artificial Proprietary Intelligence
It already exists, scattered across your systems. Not learned — discovered, resolved, governed and reasoned over in place. Customers, counterparties, decisions, outcomes, and the reasons given at the time.
The asymmetry
It cannot be bought at any price, only built. It compounds for a single organisation. And it is worth more the better the general models get.
The other race is the one nobody in the room is running.
As frontier models commoditise, the only durable differentiator left is what one organisation knows and no other can obtainThe gap
Owning it and being able to use it are different things.
As model capability rises, the advantage available from the model collapses toward zero and the advantage available from your own estate rises to almost everything. The question that then decides the outcome is a narrow one: how much of your estate can you actually put a question to?
Eleven systems, no agreement
They do not agree on what a customer is. The same organisation appears four times under four spellings, with four identifiers, and no system is wrong.
No path across them
There is no route that carries a question over the estate and returns an answer with evidence you could defend to a regulator.
Governance on a different clock
Controls sit inside each application. An agent crossing five of them is governed by none, and the record is fragmented across all five.
The route
Seven rungs — each one the reason the next is possible.
You cannot reason over what you have not resolved. You cannot resolve what you have not discovered. And you cannot govern any of it once the data has moved.
Discover
Reach the systems where they are. Types, keys, distributions, anomalies and undocumented join paths — read rather than described. No extract, no copy, no migration.
Catalogue
The estate proposes its own model — entity types, relationship types, attributes — and a person approves, corrects and constrains it. Review starts from evidence, not a blank page.
Resolve in place
Attribute matching, active learning and structural similarity decide that six records are one real entity. Thresholded, reversible, logged with its basis, with low confidence routed to a person.
Persist as a graph
Entities and typed relationships with confidence, source and date on every edge. Derived relationships that appear in no source record carry the rule that produced them and the premises it consumed.
Govern the whole graph
Inherited access-control lists are the floor — stricter, never looser — reconciled into one permission model the enterprise authors. Enforced at the traversal layer, so reasoning runs over a view that already excludes what it may not see.
Watch, do not wait
A change in the world is matched against what the enterprise knows, and the work is raised. The system stops being a place you go and becomes something that comes to you.
Act under a gate
A model proposes; a deterministic gate disposes. Write-back through the same connectors as read. Versioned, replayable, reversible, with autonomy set per task and per risk.
Every rung happens over live sources. Raw records stay behind the boundary; what persists is the resolved layer, with field-level lineage and a snapshot of each cited record so an answer survives the source moving. Nothing is lifted, nothing is copied, and no system is asked to change. That is why this is adoptable in an enterprise that cannot rip anything out — and why governance can run on the same clock as the systems it governs.
What a reached estate lets you do
None of these is a product you buy.
Each is a question the enterprise can finally ask of itself.
Relationship-aware screening
Every resolved party and its inferred connections, watched continuously rather than screened once at onboarding.
Chains to natural persons
Traversal that runs until it reaches natural persons or cannot proceed — and reports the halt, with the level and the percentage unresolved, as a finding rather than a silence.
Signal to brief
A change in a client’s world, resolved against what you already know, delivered as a brief rather than a search result.
One book after a merger
Two estates operating as one organisation, without consolidating a single system of record.
One click to a review case
Any decision promoted to a quality case, carrying its evidence, its rule versions and its full decision path.
Replay, years later
The data state, rule versions and execution path that produced a result — not an approximation of it.
Governance, made structural
Nothing enters unentitled. Nothing acts ungated.
Proof, not assurance. Every traversal, disposition and output is stamped with its source, and the whole decision can be replayed under the rules in force at the time. That is a different standard from an audit log, and it is the one a regulator, a court, an auditor or a client will actually ask for.
Entitlement filter
Every source access-control list feeds one consolidated permission model, authored and signed off by the enterprise rather than the vendor, inspectable rule by rule.
A wall around knowledge
A document wall asks whether you may open something. A knowledge wall also asks whether you may reconstruct what it would have told you.
The deterministic gate
Between a proposal and an effect. Bounded authority, contained blast radius, reversible, and a complete record of what was traversed and what was excluded.
The economics
Models are licensed. This is owned.
Every capability composed on the foundation inherits the same entities, the same permissions and the same governance. The marginal cost of the next one falls; the value of every earlier one rises as the estate resolves further.
What you license
Model access
- Converging in capability and in price
- Reaches your competitor the same quarter it reaches you
- Improves on a vendor’s schedule, not yours
- An operating cost that buys parity, never advantage
What you own
Resolved institutional context
- Customers, counterparties, decisions, outcomes, and the reasons given at the time
- Non-depreciating and non-substitutable
- Impossible to license, at any price
- Worth more the better the general models get
General intelligence is a subscription. Artificial Proprietary Intelligence is a balance-sheet item.
The first is an operating cost every competitor also pays. The second compounds for exactly one organisation, and no amount of spending elsewhere reproduces it.The operational version of the same argument: don’t buy vanilla →
Next step
Owning it is not the same as being able to use it.
The destination is not in dispute. The route is the whole question — and it runs through discovery, resolution, a governed graph and a gate, over systems that never move.