Keep the operational system. Preserve responsibility at the boundary.
Responsibility Infrastructure does not require an organisation to replace ERP, CRM, clinical, logistics, incident, case-management or custom systems simply to make responsibility more explicit and reconstructable.
Existing systems can remain authoritative for the work they already perform.
Organisations already depend on systems for records, transactions, cases, clinical activity, incidents, communications and workflow. RI is not a replacement programme for those systems. Its concern is whether responsibility meaning remains understandable when consequential work moves between people, roles, systems or organisations.
The architectural principle is additive: preserve the source system, preserve its native facts, and add only the responsibility information needed at the boundary.
RI standardises responsibility meaning, not internal technology.
A conforming implementation may use different databases, APIs, workflow engines, identity systems and user interfaces. The public requirement is that the responsibility meaning needed for interoperability and later reconstruction survives the boundary.
Detailed adapter patterns, event mappings and implementation mechanics are documented separately where they are needed for a defined implementation or interoperability purpose.
If you are evaluating how RI could sit around an existing operational platform, request the relevant implementation material for a defined system and boundary.