Skip to content
the way it works

DOCS

Concepts

The vocabulary, defined precisely. If you want the plain-language version first, start with how it works.

Glossary

Capability map
The organizing ontology. Capability → process group → process → workflow. Seven to ten level-1 capabilities, two to three levels deep, named as outcomes and mutually exclusive, so every business function has exactly one home.
Artifact
An input, never a record. A design document, procedure, runbook, spreadsheet, screenshot or diagram export. Artifacts are classified into the map and derived from; they are not the thing being stored.
Capture
The record of one process: an OCEL 2.0-aligned object-event graph. Objects are the things that move; events are what happens to them; relations between them carry qualifiers. Design-time and run-time share one schema, so a real event log imports rather than forcing a remodel.
Facet
Typed context attached to the most specific object or event it concerns. Twelve concepts: person, glossary, objective, performance, decision, requirement, assumption, policy, job-aid, system, integration, data-standard. Five of them (person, policy, performance, job-aid, system) are first-class structured data rather than free text.
Perspective
A projection of the capture, rendered on demand: object graph, process map, SIPOC, taxonomy, ledger, narrative. None of them is the record, so none of them can go stale independently.
Provenance
Four properties on every element: stated or inferred, a confidence score, the quoted evidence, and the human review state. Inferred elements render distinctly until confirmed. Nothing human-confirmed is silently overwritten. Where the source is silent, the capture is silent: empty beats invented.
Curation state
The ranking rule used everywhere the system has to choose what to surface: confirmed > reviewed > derived > auto-placed. Retrieval inherits human judgment for free.

These docs are deliberately short at launch. They will grow with the product rather than ahead of it.