Tenant is the organisation and data-isolation boundary. Project is the membership and execution scope inside a tenant. Actor is the known user, service or controlled system identity associated with an action.
Product glossary
Shared language for a connected operating model.
These definitions keep product, technical and audit conversations aligned. They describe Advanexus objects and do not redefine external regulatory or vendor terminology.
Source is a controlled connection identity. Dataset is a stable governed data contract. Version is an immutable state of a supported file, table, Dataset, Report, notebook or other versioned object.
Run is one concrete execution attempt and outcome. Artifact is a bounded retained output associated with a run. Runtime filters change execution context; server-owned RLS enforces data visibility.
Evidence is the actor, scope, change, context and outcome the canonical source can support. Finding is a supported condition requiring review. Case organises human investigation without changing the source event.
Assurance is the permission-aware read and investigation layer. Projection is its rebuildable normalised view. Integrity describes verification strength; reproducibility evaluates explicit criteria rather than an unexplained confidence score.
Scope and identity
Tenant is the organisation and data-isolation boundary. Project is the membership and execution scope inside a tenant. Actor is the known user, service or controlled system identity associated with an action.
Session registry
Server-owned record used to validate authenticated JTI, actor, tenant, active Project, lifetime and revocation state.
Permission scope
Effective operations and projects the current actor may use after server-side resolution.
Data and version
Source is a controlled connection identity. Dataset is a stable governed data contract. Version is an immutable state of a supported file, table, Dataset, Report, notebook or other versioned object.
FileVersion and TableVersion
Immutable source-file identity and immutable managed physical-table publication respectively.
DatasetVersion
Immutable Dataset-owned SQL/schema/column contract; a virtual version does not necessarily freeze live rows.
ReportVersion
Immutable analytical intent and exact DatasetVersion binding accepted for later AnalyticsRuns.
Execution and result
Run is one concrete execution attempt and outcome. Artifact is a bounded retained output associated with a run. Runtime filters change execution context; server-owned RLS enforces data visibility.
QueryExecution
One SQL Console service execution attempt with Source, SQL and success or failure context.
AnalyticsRun
One report execution with exact version bindings, filter, permission/RLS, diagnostic and artifact context.
CellRun
One ANPy code-cell execution bound to notebook revision, environment and kernel generation.
Evidence and investigation
Evidence is the actor, scope, change, context and outcome the canonical source can support. Finding is a supported condition requiring review. Case organises human investigation without changing the source event.
Evidence Package
Permission-checked bounded technical export with scope, manifest and checksum metadata; not automatically digitally signed or legally conclusive.
Service Desk Ticket
Tenant-wide operational communication record whose Project may be optional context rather than its authorisation predicate.
Assurance and integrity
Assurance is the permission-aware read and investigation layer. Projection is its rebuildable normalised view. Integrity describes verification strength; reproducibility evaluates explicit criteria rather than an unexplained confidence score.
Canonical source
Module that owns the authoritative record used by a supported workflow.
Projection
Rebuildable Assurance read model built from supported canonical events, snapshots and historical adapters.
Intelligence Turn and Action
A Turn may complete with a response; an Action records a proposed or executed registered side effect under its confirmation and approval contract.
Next step
Start with the flow that cannot afford ambiguity.
Bring the systems, owners, rules, delivery obligations and evidence requirements that matter.
Discuss a critical flow