The substrate for the AI-era enterprise.
One graph for regulations, contracts, processes, products, and people. So AI agents can reason about your business with traceability, reliability, and determinism.
The estate, not the enterprise
Walk into any large organisation and ask to see the enterprise. You will not be shown one thing. You will be shown a tour of competent modules, each authoritative for one slice. None of them is the enterprise.
DIG, the Declarative-Imperative Graph, represents the whole organisation as a single graph. Rules, regulations, obligations, contracts, processes, customers, products, controls, and the typed web of relationships that binds them. One model. Same primitives. Reasoned through one query.
01 · Declarative-imperative duality
What must be true, and what is.
Every enterprise runs on two kinds of statement. Declarative: how things must, may, or must not be. Imperative: how things are, or were. The industry has historically kept them in separate tools. DIG models them in one graph, with the same primitives, mutually informing at the granularity of a single event.
02 · Graph-native substrate
Stored as graph, queried as graph.
The questions DIG must answer are graph questions at the substrate level. Heterogeneous first-class entities, many-hop traversal in bounded work, relationships as full citizens, a single query model across structure, state, and history. Not a graph rendered from a warehouse. A graph stored and queried as graph.
03 · High-dimensional objects
Objects as citizens, not labels.
A customer, a contract, a regulation, a product, a control: each becomes a state-bearing, constraint-aware, lifecycle-governed citizen of the graph. Five dimensions per object: static, dynamic, temporal, relational, behavioural. Compliance becomes a property the object continuously evaluates about itself.
Ahead, of Complexity. of Compliance. of Change.
What it gives you
Three capabilities, one substrate.
When declarations, trajectories, and citizens live in the same graph and are queried through the same semantics, the same substrate gives you three things at once.
-
Deterministic AI
AI conclusions you can defend.
Same question, same graph state, same answer. Every answer is a path; every path is replayable; every path is auditable.
Enterprise Graph -
Continuous compliance
Conformance, evaluated at query time.
Three independent scores per subject: rules satisfied, trajectory followed, object states admissible. Live, not quarterly.
Enterprise Graph -
Live cross-domain impact
When something changes, see everything else that changes.
A change in one domain propagates along typed dependencies to every trajectory, declaration, or object whose meaning depends on it.
Enterprise Graph
Read the paper
DIG: a universal modelling methodology for the AI era.
A 60-page methodology paper. The three moves, four reasoning classes, comparisons to adjacent software categories. Draft v4, on this site.