- The 10 Data Domains Every EA 2.0 Graph Needs
- 1 Purpose
- 2 Design Philosophy
- 3 The Ten Core Data Domains
- 4 Relationship Map
- 5 Data Health Attributes
- 6 Integration Sequencing
- 7 Source Stewardship
- 8 Sovereign Cloud Considerations
- 9 Minimum Viable Graph Readiness Checklist
- 10 Future Data Domains (Optional)
- 💡 Takeaway
The 10 Data Domains Every EA 2.0 Graph Needs #
1 Purpose #
A graph is only as intelligent as the data that feeds it.
EA 2.0 succeeds not by building new data, but by integrating the living systems already running the enterprise — each a view into capability, cost, risk, or compliance.
The goal of this inventory is to define what must be connected for an EA 2.0 platform to generate predictive, auditable insight.
2 Design Philosophy #
- Breadth first → Depth later: connect all 10 domains lightly before enriching any one of them deeply.
- Native source → Authoritative system: always extract from the system of record, not a shadow report.
- Metadata over payload: EA 2.0 cares more about “what, who, when, why” than raw data volume.
- Low-impact collection: never stress production; use deltas, APIs, or replicas.
- Traceability: every record in the graph must know its origin (
source_system,source_id,last_seen_at).
3 The Ten Core Data Domains #
| # | Domain | Source Systems (typical) | Key Entities to Extract | Why It Matters to EA 2.0 |
|---|---|---|---|---|
| 1 | Business Capability Register | Excel catalog / SharePoint / Collibra / Strategy Tool | Capabilities, owners, KPIs, strategic themes | Defines the anchor layer of the graph; links business intent to delivery. |
| 2 | Application Portfolio / CMDB | ServiceNow CMDB / BMC / LeanIX | Applications, services, interfaces, lifecycles | Core visibility of “what runs where”; enables rationalization and impact analysis. |
| 3 | Infrastructure & Cloud Inventory | Azure Resource Graph / AWS Config / vCenter | Compute, storage, network assets, tags | Provides run-time context and enables cost and risk analytics. |
| 4 | Data Catalog & Lineage | Purview / Glue / Informatica | Datasets, schemas, sensitivity, lineage links | Connects data flows to apps and controls for trust and AI readiness. |
| 5 | Risk & Control Registers | ServiceNow GRC / Archer / IRM | Risks, controls, mitigation plans | Powers predictive governance and policy simulation. |
| 6 | Project & Investment Portfolio | PPM Tool / Azure DevOps Boards / Jira | Initiatives, budgets, timelines, objectives | Maps strategic spend to capability change; tracks value realization. |
| 7 | Finance & Procurement Feeds | ERP / SAP / Excel reports | Application costs, contracts, vendors | Enables cost forecasting and vendor risk profiling. |
| 8 | Security Posture & Compliance | Defender for Cloud / Prisma / Security Hub | Vulnerabilities, compliance status, severity | Feeds AI models for risk propagation and self-healing actions. |
| 9 | HR & Org Structure | Workday / SAP SuccessFactors | People, roles, departments, tenure | Establishes ownership and accountability graphs. |
| 10 | Observability & Incident Telemetry | ITSM / APM / Log Analytics / Sentinel | Alerts, MTTR, incident tags | Closes the loop — feeds AI learning and governance feedback. |
4 Relationship Map #
Each domain links to others in predictable ways:
[Capability]
↕ supports
[Application] → uses → [Data Asset]
↕ hosts
[Infrastructure]
↕ governed by
[Control] ← mitigates ← [Risk]
↕ owned by
[Role]
↕ measured by
[Outcome]
The goal isn’t completeness — it’s minimum viable connectivity: every node can reach a business outcome in ≤ 4 hops.
5 Data Health Attributes #
For every connected domain, EA 2.0 tracks:
| Attribute | Description | Example Metric |
|---|---|---|
| Freshness | Last update age | 95 % ≤ 30 days |
| Coverage | % of apps linked to capabilities | ≥ 90 % |
| Completeness | Populated required fields | ≥ 85 % |
| Confidence Score | Composite trust index (0–1) | 0.8+ target |
| Sensitivity | Highest label in dataset | Confidential / Restricted |
6 Integration Sequencing #
- Quick-Win Stage – Connect CMDB, Capability Register, and Data Catalog (foundational graph).
- Expansion Stage – Add Risk/Control, Finance, and Project Portfolios.
- Optimization Stage – Bring Security Telemetry + HR Ownership for predictive governance.
- Closed-Loop Stage – Integrate Observability and Incident data for self-healing triggers.
7 Source Stewardship #
Each domain has a data steward responsible for freshness and quality.
EA 2.0 automates reminders and dashboards for them:
- Red flags when data hasn’t refreshed within SLA.
- Confidence scores per stewardship group.
- Auto-assignment of correction tasks via ServiceNow.
8 Sovereign Cloud Considerations #
- All connectors run within the tenant’s sovereign network boundary.
- API calls authenticated via Entra ID (OAuth 2.0 client credentials).
- Sensitive feeds (e.g., HR or Finance) are tagged
restrictedand never leave region. - AI reasoning services query anonymized metadata only.
9 Minimum Viable Graph Readiness Checklist #
✅ Capability → Application links ≥ 80 %
✅ Application → Data → Control chains exist for critical assets
✅ At least one Risk node connected to each Capability domain
✅ Cost data available for top 20 apps
✅ All nodes have source_system and last_seen_at
✅ Feed frequency ≤ 30 days
Meeting these marks unlocks the Predictive and Governance layers.
10 Future Data Domains (Optional) #
- Customer Experience Metrics (NPS, churn predictors)
- Sustainability Data (Carbon intensity per service)
- Third-party Dependency Maps (SaaS integration risk)
- AI Model Registry (Responsible AI metadata)
These extend EA 2.0 into strategy, ESG, and AI governance territory.
💡 Takeaway #
The EA 2.0 graph is not built once — it is fed.
A living architecture is a living integration network.
When these ten domains stream in sync, the organization finally sees itself as one system.
