Memory¶
Memory is the capability to retain organizational understanding beyond individuals, applications, projects, and technology changes.
In the Organizational Intelligence Framework, this capability is called Organizational Memory.
Role in the Organizational Intelligence Cycle¶
Memory supports the Remember stage of the Organizational Intelligence Cycle.
It allows the organization to carry useful experience forward so future reasoning, decisions, actions, and learning can improve.
Without memory, the organization must repeatedly reconstruct context and may make similar decisions without benefiting from prior experience.
Responsibilities¶
The memory capability should:
- preserve relevant observations, context, decisions, rationale, outcomes, and lessons
- connect memory records to source evidence
- distinguish current knowledge from historical records
- support retrieval by future reasoning and workflows
- retain context across people, systems, vendors, and organizational changes
- govern retention, deletion, access, and sensitivity
- support portability where practical
- maintain traceability from memory to decisions and outcomes
Inputs¶
Possible inputs include:
- interpreted observations
- contextual relationships
- decisions and rationale
- policy references
- evidence records
- action records
- outcome measurements
- lessons learned
- review and audit findings
- model or automation evaluation results
Outputs¶
Outputs may include:
- reusable context
- decision history
- evidence links
- lessons learned
- prior-case references
- policy interpretation history
- outcome summaries
- memory retrieval results
- records for audit and governance
Controls¶
Memory must be strongly governed because it may contain sensitive, regulated, or business-critical information.
Controls should address:
- access control
- data classification
- retention and deletion rules
- consent and privacy requirements
- source provenance
- lineage and versioning
- audit logs
- legal hold requirements
- review of sensitive or high-impact memory records
- portability and export requirements
Quality Measures¶
Memory quality can be assessed through:
- completeness of required context
- traceability to source evidence
- retrieval relevance
- freshness and currency
- retention compliance
- duplicate and contradiction rates
- explainability of decision history
- coverage of critical decision domains
- reuse rate in future reasoning and workflows
High memory quality means the organization can find, trust, and reuse prior experience when making future decisions.
Anti-patterns¶
Common anti-patterns include:
- storing documents without context
- treating archives as Organizational Memory
- preserving facts but losing decision rationale
- retaining sensitive information without governance
- relying on individual memory for repeated decisions
- locking important memory inside one tool or vendor
- failing to link outcomes back to decisions
- deleting context needed for future accountability
- keeping outdated memory without versioning or status
Implementation-neutral Examples¶
Examples of memory include:
- a record of why a policy exception was approved, including evidence, approver, rationale, date, and later outcome
- a reusable history of similar customer issues and their resolutions
- a decision log linking recommendations, human review, action taken, and result
- retained context about a document's source, owner, effective period, related process, and superseded versions
- lessons from a fraud investigation connected to future risk review guidance
These examples do not require a specific database, knowledge graph, vector store, content management system, or AI model. Those may be implementation choices, but the framework requirement is the organizational capability.
Unless otherwise noted, this document is licensed under CC BY 4.0.