advanexus

advanexus Living System Map

How advanexus works

From source and execution to an evidence-backed business outcome.

Progress that stays connected.

advanexus brings advance and nexus together: moving forward without breaking the connection between data, decisions, accountability and evidence.

Depth
  1. L0
  2. L1
  3. L2
  4. L3
  5. L4
  6. L5

All records and identifiers shown here are synthetic examples.

63 Interactive advanexus system map

    Find a module or object

    Find a module or object

    Play a business story

    Play a business story

    Text view

    The system as a controlled flow

    Seven operating corridors and a cross-cutting trust layer connect intent to an evidence-backed outcome.

    1. 01 Connect Start with a business need, a connection identity or an immutable FileVersion; connector capabilities remain explicit.
    2. 02 Prepare Validate SQL and file inputs, and preserve QueryExecution or TableVersion as distinct, inspectable records.
    3. 03 Execute Run versioned pipeline intent through explicit PipelineRun and QualityRun records; a stored schedule is not guaranteed dispatch.
    4. 04 Govern Make DatasetVersion the accepted contract for source, SQL, schema and lineage without claiming a physical row snapshot.
    5. 05 Analyse Separate mutable Report Draft, immutable ReportVersion, concrete AnalyticsRun and the visualization used to read its result.
    6. 06 Operate Turn governed results into human decisions, dashboards, Python analysis and confirmed canonical actions.
    7. 07 Prove Follow findings, cases, tickets, integrity checks and evidence packages without turning Assurance into a second source of truth.
    8. 08 Trust Recheck Tenant, Project, Actor, permissions and RLS at every operation; audit projections may arrive eventually.

    Canonical object index

    • Business need / obligation decision Requirement โ†’ Approved intent
    • Source object Connection identity โ†’ Capability contract
    • FileVersion version File / secure URL โ†’ Immutable file reference
    • SQL Console module Source + SQL โ†’ QueryExecution
    • QueryExecution run Validated query โ†’ Result / lineage
    • SavedQuery definition QueryExecution + SQL intent โ†’ Reusable query definition
    • System Sandbox Source object FileVersion โ†’ TableVersion
    • TableVersion version FileVersion + bounded transform โ†’ Managed table revision
    • Pipeline / JobVersion versioned definition JobDefinition + immutable JobVersion โ†’ Pinned PipelineRun
    • PipelineRun run Pinned JobVersion + effective policy + admission contract โ†’ Attempts ยท state events ยท checkpoints ยท Master steps
    • QualityRule definition Source + expectation โ†’ QualityRun
    • QualityRun gate QualityRule + Source โ†’ Pass / fail + Finding
    • QualityFinding finding Failed persisted QualityRun โ†’ Bounded failed-run evidence
    • Dataset identity Accepted definition โ†’ DatasetVersion
    • DatasetVersion contract Source ยท SQL ยท schema โ†’ Accepted contract
    • Report identity Name + analytical purpose โ†’ Drafts + immutable versions
    • Report Draft draft Mutable Report asset โ†’ Non-runnable draft
    • VisualizationDraft transient preview Actor-owned QueryExecution โ†’ Bounded preview
    • ReportVersion version DatasetVersion or SQL snapshot โ†’ Immutable report intent
    • AnalyticsRun run ReportVersion + filters โ†’ Result + diagnostics
    • ANPy workspace ยท NotebookVersion versioned workspace Immutable READY Environment + saved cell source โ†’ CellRun
    • CellRun run NotebookVersion + saved cell + EnvironmentVersion โ†’ Bounded output + diagnostics + evidence
    • Visualization lens Analytics result โ†’ Human-readable lens
    • VisualizationWorkspaceVersion immutable composition Exact ReportVersions + responsive layout โ†’ DashboardRun + version-pinned controls
    • DashboardRun composite run Published VisualizationWorkspaceVersion + actor filters โ†’ Distinct WidgetRun outcomes
    • WidgetRun run Exact workspace binding + ReportVersion + filters โ†’ Actor-visible AnalyticsRun
    • Business outcome outcome Supported result โ†’ Decision / delivery
    • Finding object Supported signal โ†’ Owned issue
    • Assurance Case remediation Finding โ†’ Remediation trail
    • Service Desk Ticket operational workflow Support request โ†’ Messages + status events
    • Entity 360 / Execution Story read model Supported source events + audit evidence + bounded adapters โ†’ Permission-filtered source-aware story + visible gaps
    • IntegrityRun run Permitted evidence references + explicit criteria โ†’ Criterion results + visible gaps
    • EvidencePackage artifact Permitted evidence set โ†’ Manifest + checksums
    • Intelligence proposal proposal Goal + permitted context โ†’ Bounded proposal
    • Validation + confirmation gate Proposal โ†’ Authorised action
    • Registered Action run Confirmed request โ†’ Result + evidence
    • Tenant ยท Project ยท Actor scope envelope Identity + membership โ†’ Bounded scope
    • Permission ยท RLS ยท Policy control envelope Actor + target โ†’ Effective access context
    • Audit ยท Outbox ยท Correlation evidence bridge Module-specific state change โ†’ Source-specific projectable evidence

    Read the business stories

    1. Regulatory reporting Follow the controlled path: Business need / obligation โ†’ Source โ†’ SQL Console โ†’ QueryExecution โ†’ Dataset โ†’ DatasetVersion โ†’ ReportVersion โ†’ AnalyticsRun โ†’ Visualization โ†’ Business outcome โ†’ Entity 360 / Execution Story โ†’ IntegrityRun โ†’ EvidencePackage.
    2. Controlled migration Follow the controlled path: Business need / obligation โ†’ FileVersion โ†’ System Sandbox Source โ†’ TableVersion โ†’ Pipeline / JobVersion โ†’ PipelineRun โ†’ QualityRun โ†’ Dataset โ†’ DatasetVersion โ†’ Entity 360 / Execution Story.
    3. Governed self-service analytics Follow the controlled path: Source โ†’ SQL Console โ†’ QueryExecution โ†’ Dataset โ†’ DatasetVersion โ†’ Report โ†’ ReportVersion โ†’ AnalyticsRun โ†’ Visualization.
    4. Data Quality blocks incorrect delivery Follow the controlled path: Source โ†’ Pipeline / JobVersion โ†’ PipelineRun โ†’ QualityRun โ†’ QualityFinding โ†’ Audit ยท Outbox ยท Correlation โ†’ Entity 360 / Execution Story โ†’ IntegrityRun.
    5. Incident and recovery Follow the controlled path: Business outcome โ†’ Finding โ†’ Assurance Case โ†’ Pipeline / JobVersion โ†’ PipelineRun โ†’ QualityRun โ†’ Dataset โ†’ DatasetVersion โ†’ Entity 360 / Execution Story โ†’ IntegrityRun โ†’ EvidencePackage.
    6. AI-assisted controlled action Follow the controlled path: Business need / obligation โ†’ Intelligence proposal โ†’ Validation + confirmation โ†’ Registered Action โ†’ Audit ยท Outbox ยท Correlation โ†’ Entity 360 / Execution Story.
    7. File to governed data Follow the controlled path: FileVersion โ†’ System Sandbox Source โ†’ TableVersion โ†’ Dataset โ†’ DatasetVersion โ†’ ReportVersion โ†’ AnalyticsRun.
    8. Evolve a versioned contract Follow the controlled path: Dataset โ†’ DatasetVersion โ†’ Report โ†’ ReportVersion โ†’ VisualizationWorkspaceVersion โ†’ DashboardRun โ†’ WidgetRun โ†’ AnalyticsRun โ†’ Visualization โ†’ Business outcome โ†’ Entity 360 / Execution Story.
    9. Controlled Python analysis Follow the controlled path: Business need / obligation โ†’ ANPy workspace ยท NotebookVersion โ†’ CellRun โ†’ Audit ยท Outbox ยท Correlation โ†’ Entity 360 / Execution Story โ†’ IntegrityRun.
    10. Service Desk hand-off Follow the controlled path: Service Desk Ticket โ†’ Assurance Case โ†’ Entity 360 / Execution Story โ†’ IntegrityRun โ†’ EvidencePackage.

    Product boundaries

    • The map can expose supported facts and evidence; the business obligation, approval and final decision remain human-owned.
    • Metadata, preview, query, transfer, write, delete and quality capabilities vary by connector, driver and accepted deployment.
    • A stable identity or mutable definition is not a run. An explicit version reference preserves the accepted intent.
    • A query is an operation or mutable working definition, not a governed Dataset identity.
    • Definitions and versions do not prove execution. Each run is a separate record with its own scope, status and outcome.
    • A future schedule occurrence performs fresh admission checks and pins the then-current immutable JobVersion in a durable PipelineRun before dispatch; exactly-once business effects and rollback depend on the executor, connector, policy and proven idempotency.
    • QualityRule is currently a mutable source contract, not an immutable rule pinned to DatasetVersion; QualityRun remains separate from PipelineRun.
    • DatasetVersion locks the accepted contract, SQL, schema and relationships, not physical source rows. Row reproduction needs a retained result, artifact or materialised snapshot.
    • A draft is not executable. ReportVersion may be pinned to DatasetVersion or an immutable SQL snapshot; runtime filters belong to AnalyticsRun and do not mutate it.
    • VisualizationDraft is a transient, actor-owned view of one bounded QueryExecution snapshot and does not rerun SQL. Saving creates canonical Dataset and Report versions; the draft itself remains non-authoritative.
    • Visualization is a human-readable lens over a result, not a data source and not proof that the result is correct.
    • A published VisualizationWorkspaceVersion freezes its responsive layout and exact ReportVersion and DatasetVersion bindings. DashboardRun โ†’ WidgetRun โ†’ actor-visible AnalyticsRun preserves every outcome, including partial success. Refresh schedules, prepared data, alerts and controlled sharing are separate version-pinned, re-authorised controls; sharing and export remain deployment-policy gated.
    • Assurance is a permission-aware read model over canonical sources. It does not own, rewrite or silently complete source execution history.
    • A Service Desk Ticket is an operational support workflow, not an Assurance Case and not proof that remediation succeeded.
    • EvidencePackage is a technical export with a manifest and checksum metadata; it is not automatically digitally signed, WORM or a legally final record.
    • An Intelligence proposal requires deterministic validation, human or policy confirmation and fresh authorisation before a canonical service performs the action.
    • Tenant, Project, Actor, membership, permissions and RLS are rechecked for each operation. A role lens is not an authorisation system.
    • Audit and outbox projections are eventually consistent read-side evidence. Canonical service state remains authoritative and missing evidence stays visible.