Coherence Compute Engine

Connections are computed the same way every time.

Computable ontology.

An ontology used to be drawn by hand, and the schema grew with every project that touched it. Past a certain size it was too large and too irregular to use, and too irregular for software to read at all. It aged faster than anyone could correct it.

Two things changed. The set of types is fixed: six zones, 24 kinds of thing, 31 kinds of connection, identical in every deployment, so every ontology comes out the same shape. And the reading is done by machine, at the speed your material arrives: the part that used to be manual work nobody sustained.

That is what makes the ontology computable. Checks run over it the same way every time, and software can depend on the output. The AI reads your material and proposes what it found. It does not decide what connects to what, and it does not decide what is true.

Nobody opens this directly. It is the compute under the other three.

From your material to a signed result.

ReadsDocuments, procedures and registers; schema exports, code and interfaces where you connect them. Structure only.
ComputesEvery connection scored against the fixed set of types, the same way every time. The same material always yields the same links, and no language model decides what connects to what.
GradesCertain links stand on their own. Uncertain ones wait in a separate list of proposals for your experts, and the two counts are reported separately.
Names the basisEvery connection records what it rests on: a parse of the source, a person’s confirmation, or the machine’s proposal at a measured threshold. The label never upgrades, so a machine’s guess cannot read back as a person’s judgment.
Declares absenceThe fixed types say what should exist, so a missing document or control is computed as a finding with a name.
Stays in your perimeterReading and linking run on models inside your perimeter. Your material is not sent out to be read, and there is no per-document call to an external model.
SignsEvery result carries the path it took and the evidence behind it. 1,957 deterministic tests guard the behaviour that must never vary.

A fixed set of types, so software can depend on the output.

Six zones. 24 kinds of thing. 31 kinds of connection. That is the entire vocabulary, locked in code. The platform reads your material and proposes; your experts accept, correct or reject. Nothing probabilistic decides what is true.

Motivation

Why it exists, what it is for

Strategy

Where it is going

Capability

What the business can do

Operations

How the work actually runs

Governance

Rules, controls, sign-off

Risk

Exposure and failure paths

No customer-editable metamodel.

Locked in code, identical in every deployment.

No customer-editable schemaEvery deployment runs the same set. A schema each customer extends produces models that cannot be compared and cannot be checked by the same rules, which is how modelling tools stop being used.
What is unboundedYour own terms, inside the fixed set. The types are fixed; what fills them is your business, without limit. Complexity goes into your content, not into the schema.
How new needs enterAs separate dimensions alongside the fixed set: process order, infrastructure, cross-system mappings, who is accountable for what. The set itself does not move.

The seven rules it checks.

Fixed like the types, run on every model. Each finding names the rule that fired and carries its severity.

Regulation without a control

An obligation on the books with nothing enforcing it. Critical.

Risk without a control

A named risk with no control, policy or process mitigating it. Critical.

Goal without an initiative

A stated goal that nothing is funded to reach. High.

Initiative without a capability

An initiative with no capability, process or system to execute it. High.

Capability without a system

Work the business does with nothing automating it. Medium.

System without a capability

A system that no capability claims. Medium.

Stakeholder without a goal

A role with no goal, driver or initiative attached to it. Medium.

The other three products.

All four read and write the same model. See the whole platform.

Coherence Ontology Studio

You scope the domains. The platform reads your material and proposes what it found, and what your experts confirm is the model.

Coherence Explorer

Ask it, and the answer names its sources. Test a change and see what it touches before you commit.

Coherence Change Orchestrator

Any initiative modelled on the transformation map, then driven through two gates to the work packages that land it.

See a worked example.

Three are open to explore, or start your own first domain with five documents.

This site uses cookies

We use essential cookies for the site to function and analytics cookies (Google Analytics) to understand how you use it. Analytics cookies are only activated with your consent. We do not track you across other websites. Your data is stored in the EU and processed in accordance with GDPR. Read our Privacy Policy