Shared Memory
Institutional memory should describe the world—not the filing system.
Shared Memory organises evidence around persistent entities and their histories. New information updates the institution's understanding while preserving source lineage, time, authority and prior states.
From source to memory
Turn every mandate-relevant source into structured institutional understanding.
The ingestion layer identifies meaningful units inside each source, links them to the correct entities, records when they were true or observed and preserves where they came from. A Data Block can represent a fact, event, observation, relationship, evidence item, change or assertion — structured enough to be queried and reused, yet always connected to its source.
Memory is extended only for authorised sources and only within the scope approved for each delivery. Documents, records, workflow history, policy rules and entity knowledge grow release by release, so shared memory always reflects agreed, governed coverage rather than everything the institution holds.
Source examples
- Reports
- Case records
- Emails and correspondence
- Authorised Viber and online messaging records
- Forms and spreadsheets
- Media and evidence files
- Mandate and authority changes
- Executive Director directives
- RA 11930 obligations and Strategic Plan requirements
- Decisions and action confirmations
- Institutional and international updates
Provenance
Where the information came from.
Time
When it occurred, applied or was observed.
Entities
Who or what the information concerns.
Meaning
The defined institutional semantics.
Authority
Which mandate or rule makes it relevant.
Access
Who may see or use it.
Status
Whether it is asserted, verified, superseded, disputed or under review.
Entity-centred memory
Memory accumulates around the things the institution is responsible for understanding.
Each entity has a time-aware history assembled from authorised Data Blocks. This allows the system to answer not only “what do we know now?” but also “what did we know at the time, why did we believe it and which evidence changed the understanding?”
Entity examples
- People and roles
- Agencies and institutions
- Countries and jurisdictions
- Mandates and authorities
- Executive Director directives and assignments
- Cases and matters
- Issues and risks
- Obligations and requirements
- Statutory and Strategic Plan compliance indicators
- Decisions and actions
- Policies and operating rules
- Outcomes and exceptions
Rich Data Block
A Rich Data Block turns institutional analysis into a reusable information product.
A Rich Data Block assembles relevant Data Blocks and entity memory into a governed conclusion, analysis or outcome. It may be rendered as a report, advice, decision, referral, coordination package or structured response.
Its meaning remains stable when it moves to the next authorised participant. Its conclusions remain linked to the underlying evidence. Its structured content remains queryable even after the document has been issued.
Registers
Registers become views of memory, not isolated sources of truth.
Existing registers remain useful operational views. The difference is that they can be generated from a shared, governed memory rather than maintained as independent stores that drift apart. Different users can see the view they need while the underlying entities, evidence and semantics remain consistent.
Permission-aware by design
Shared does not mean unrestricted.
Every query, context package and agent action must respect institutional authority, role, purpose and information sensitivity. The same entity memory can support different authorised views without exposing every underlying Data Block.