Skip to main content

Glossary

Home · Maintenance, support and FAQ · Reports, AI Log and traceability

A

Governed action

Operation prepared in the application then submitted for approval before execution, e.g., a SharePoint publication or ticket creation.

Agent

Specialized conversational component that works in the context of the current project and can produce a structured output, citations and recommendations.

Approvals / approvals

Validation step used to accept, reject or frame an action before its execution.

Artifact

Intermediary or final governed object usable in Reports & artifacts.

Artifact destination / destination of artifact

Governed target used during publication or document distribution, e.g., SharePoint.

B

Project binding

Project‑level link that makes a platform integration actually usable in a given project.

C

Portfolio command center

Multi‑project comparison surface based on signals, weights, thresholds and saved cohorts.

Citations

References displayed with certain agent results or runs to show evidence or sources used.

Cohort

Saved set of projects and portfolio comparison parameters.

Confidence

Indication attached to a structured output to help judge the solidity of the result.

Knowledge

Space where reference documents are searched, uploaded, imported and filtered for the project.

Context snapshot

Capture of the context associated with a run in the AI Log.

D

Deployment‑selected AI Provider / AI provider selected at deployment

Provider chosen during installation or deployment of the environment.

Diff

Comparison between two artifact versions to review differences before approval or publication.

Digest

Governed summary of several signals, visible in Workspace and can serve as a basis for a notification draft. See also the page Governance, decisions and actions.

Document categories / documentary categories

Classification used to organize project contents and improve their exploitation.

E

Effective AI Provider / effective AI provider

Provider actually used for a given run. This is the reference value in the AI Log.

Entitlement

Commercial or technical capacity authorized by the plan, can unlock or block a connector, provider or premium feature.

Activity event

Entry in the Activity timeline in the AI Log, with type, kind, actor and raw payload.

Workspace

Central page of the project gathering summary, operational transparency, signals and project‑level settings.

Execution connector / execution connector

Governed connector used to launch an external action, e.g., Teams, Jira or SharePoint publish.

Executed

State indicating that a governed action has actually been launched and logged, beyond just approval.

F

Model family

Transparency information visible in the AI Log to indicate the type of model actually used.

Fresh / aging / stale / conflicting / unavailable

Freshness states used to qualify evidence or source: fresh, aging, stale, conflicting or unavailable.

Ingestion provider

Technical source allowing content import into Knowledge.

H

Health / health status

Availability or state to check applied to a connector, provider or global health panel.

Held

State of holding a draft or distribution flow that should not yet leave its target channel.

J

AI Log

Page that exposes runs and associated activity for AI treatments, with trace ID, context, citations and related artifacts.

L

Read‑only

State where a page or tab remains visible but not editable.

Lineage

Traceability chain linking a run, structured output, artifact, PM Doc and associated evidence.

LLM deployment name

Exact name of LLM deployment expected by some providers, especially Azure OpenAI, in platform administration.

O

Operational

Status indicating that a provider can be considered exploitable after configuration, validation and successful test.

P

Command palette

Quick access mechanism to pages and common actions from the top bar.

Pending approval

State of a governed request or action awaiting human approval before execution.

PM Doc

Project management document revised, edited, downloaded, published or re‑injected into knowledge from Reports & artifacts.

Render profile

Parameter used to determine the format or rendering behavior of generated document.

R

Reports & artifacts

Surface used to review, compare, edit, publish and download deliverables.

Readiness

Level of readiness of a project, integration or governance flow to work without operational blockage. See also Portfolio and technical administration.

Run

Recorded execution of an AI or agent exchange, accessible later in the AI Log.

S

Seat

License unit consumed by a user in deployments where access is limited by subscription.

Signal

Proactive element that raises attention, e.g., contradiction, blockage, low freshness or follow‑up.

Source label

Label of provenance that helps identify a readable source.

Source system

Source system displayed with evidence or result.

Structured output

Formatted result of an agent or run, possibly including summary, decisions, citations, confidence and metadata.

T

Dashboard

Orientation page that exposes general state, metrics and next suggested actions.

Test

Connectivity test executed after validation of a provider.

Trace ID

Technical identifier of a run or event, indispensable for investigation and support.

Validate

Administrative step that verifies consistency before the test of connectivity or activation of a provider.

V

Validate / validation

Administrative step that checks consistency.

Artifact version

Historical state of an artifact, used for diff, lineage, approval and publication. See also Reports, AI Log and traceability.

W

Watchpoint

Explicit point of attention in a structured output, often used to signal a risk, contradiction or zone to monitor.

Next