The Data Integration and Interoperability Plan: Contracts, Flows, and Standards
Most organizations didn't design their integration landscape â it accumulated. A file export here, an API call there, a nightly batch job someone built years ago. Over time, the result is a fragile web of point-to-point connections that nobody fully understands, where a change in one system unexpectedly breaks three others.
The Data Integration & Interoperability Plan â aligned with the data integration and interoperability knowledge area of DAMA's DMBOK â brings order to that web. This article explains its contents and how ADPA generates it.
What the Plan Contains
| Section | Purpose |
|---|---|
| 1. Purpose and scope | Which systems and data flows are covered |
| 2. Integration inventory | All current integrations between systems |
| 3. Contract register | Formal integration contracts between pairs of systems |
| 4. Data flow mappings | Which data moves from which source to which target |
| 5. Interoperability standards | Agreed formats, protocols, and semantics |
| 6. API inventory | Available APIs and interfaces |
| 7. Error handling and retries | How failures are detected, retried, and escalated |
| 8. Change management | How integration changes are agreed and communicated |
| 9. Review cadence | When the plan is reviewed |
The contract register is the core idea: every integration is treated as an explicit agreement between two systems â what data, in what format, how often, with what quality expectations, and who owns each side.
How ADPA Generates It
Inputs
- Systems of record and integration patterns from the Enterprise Data Architecture Charter.
- Data domains from the Data Governance Charter.
- The existing API and interface inventory, and existing integration documentation.
Grounding
Before drafting, ADPA queries its Governance Knowledge Graph for:
- Designated systems of record and approved patterns, so every contract references real systems and approved ways of integrating.
- Existing integration contracts, so the draft doesn't create duplicate or conflicting contracts for the same system pair.
Outputs
Structured entities for integration contracts, data flow mappings, interoperability standards, and API inventory entries â which the Metadata Management Plan later uses to build lineage.
Quality Gates
- System traceability: every contract names two systems already designated as systems of record (blocking).
- No dangling flows: every data flow mapping has both a source and a target contract (blocking).
- Pattern consistency: each contract's integration pattern matches the pattern defined for that domain in the architecture charter (warning).
Where It Fits
This plan turns the architecture charter's integration patterns into concrete, governed contracts. Its contracts and flow mappings are the raw material for data lineage in the Metadata Management Plan, and they underpin reliable feeds into master data and BI.
Conclusion
Integrations are where data quality problems, outages, and security gaps often begin. Treating each integration as an explicit contract between known systems, using approved patterns and shared standards, turns an accidental web of connections into a managed, understandable landscape.