End-to-end governance in a space

Use DAM AI conversations for individual tasks. Create a DIG data intelligence space when you need to govern multiple objects, track changes, or retain policies and results as reusable context. The space connects issue discovery, inputs, execution, validation, and publishing in a traceable workspace.

This guide describes a lifecycle, not a fixed built-in automation. Available agents, Sprites, Routines, graphs, and publishing options depend on license, configuration, and permissions.

Choose a workspace

Task	Entry point	Result handling
Single table, field, or one-time question	DAM AI conversation	Review proposals and verify the relevant DAM module
Multiple object types and documents in a domain	DIG space	Track assets, documents, execution, and acceptance together
Repeated governance or metadata change monitoring	Space Routines and schedules	Review and verify triggered results according to configuration
Context queries from external AI applications	Context Management and Publish	Validate before publishing a space Agent or MCP

Step 1: Assess governance

1. Open DIG directly or use Switch to DIG in DAM. Use Switch to DAM to return.
2. Ask AI to assess metadata coverage, business semantic completeness, standards associations, lineage, and classification and security levels.
3. Define a goal, business scope, acceptance criteria, and owner.
4. Query candidate data sources, tables, fields, reports, or other objects and add confirmed objects to the governance inventory.

The assessment proposes scope and priorities; it does not complete governance. Verify each object's source, system, schema, and permissions before adding it.

Step 2: Create the space

1. Create a space named for the business domain and goal. Document scope, owner, and acceptance criteria.
2. Add inventory objects as governance inputs.
3. Add requirements, standards, naming rules, business policies, metric definitions, and acceptance documentation.
4. Before Data Governance, verify sources, effective document revisions, and access scope.

Assets and documents provide context without automatically copying source data or replacing source system, DAM, or document permissions. See space operations.

Step 3: Run governance Routines

Select an appropriate workflow on the right and specify objects and reference documents. Routines combine Sprites to enrich metadata, map standards, generate quality rules, classify data, or organize assets.

• Run now: suitable for initial validation and small, observed tasks. Start with a sample.
• Scheduled execution: configure frequency, scope, failure handling, and notification owner before enabling recurring runs.
• Proactive governance: monitor metadata changes and trigger configured tasks. A trigger does not automatically approve or publish proposals.

Check inputs, Routine revision, object count, and reviewer before execution. Distinguish successful, failed, skipped, and pending-review results afterward. See Sprites and Routines and Routine operations.

Step 4: Track progress

Capability	Purpose	Verify
Space AI conversation	Ask about status, gaps, and next steps	Answers reference current objects and valid documents
Governance graph	Explore assets, standards, quality, security, and task relationships	Relationships reflect confirmed results
Asset profile	Summarize semantics, ownership, quality, and usage	Gaps are resolved in the appropriate DAM module
Business lineage	Understand sources, flows, and impact	Lineage has been harvested, parsed, and reviewed

Cross-check these views rather than relying on one progress percentage. Resolve missing objects, incomplete semantics, and unreliable relationships in the relevant DAM module before publishing.

Step 5: Review results and scorecards

Review execution, pending approvals, and governance scorecards against organizational targets. A scorecard summarizes completion and gaps; it is not a single verdict on business data quality.

1. Confirm that every inventory object has an outcome, including reasons for failure or skipping.
2. Complete required review and verify results in DAM.
3. Check that semantics, standards, lineage, quality, and security support the target use case.
4. Sample space AI answers. Verify references and that unsupported or out-of-scope questions are identified as such.
5. Assign remaining gaps to the next scope, owner, and acceptance date.

Proceed only after confirming results, context quality, and access scope. Execution success or a score threshold does not replace business acceptance.

Step 6: Publish knowledge for AI

In Context Management, organize confirmed semantics, standards, lineage, and reviewed documents, then open Publish.

1. Select context and exclude unrelated, expired, or unreviewed content.
2. Validate typical answers, citations, and access boundaries using the query agent.
3. Publish the space knowledge base as MCP for external AI applications, or provide a space Agent.
4. Configure consumers, authorization scope, and client connection details; record the revision.
5. Recheck context and published capabilities when objects, standards, permissions, or documents change.

Space MCP supplies governed semantics, knowledge, and relationships; it does not directly open underlying business data. Inbound space publishing and DIG's outbound MCP calls are separate configurations. See context publishing.

Acceptance criteria

• Object sources, scope, assets, and documents are traceable.
• Executions, schedules, and change triggers have records; unresolved items have outcomes.
• Key results are reviewed and verified in DAM.
• Graphs, profiles, lineage, and scorecards reflect confirmed results.
• Published Agent or MCP tools contain only checked, authorized context.
• Owners and plans cover subsequent governance, validation, and republication.

Related guides

• DIG overview
• Quick start
• Data intelligence spaces
• Governance capabilities
• Context publishing
• Approvals