Skip to content

Learn

The vocabulary of the category.

These are the terms of Organizational Memory: what they mean, why they matter, and where they come from.

01 Glossary

The vocabulary of Organizational Memory

  1. Term 01Organizational MemoryOrganizational Memory is the persistent, shared record of what an organization knows: its decisions, commitments, processes, and relationships, with the reasoning behind them, held as infrastructure the organization owns rather than scattered across individual tools, inboxes, and employees' heads.
  2. Term 02Organizational Memory LayerThe Organizational Memory Layer is the infrastructure tier that maintains an organization's memory beneath every business function and every AI system, so context accumulates once and is inherited everywhere instead of being rebuilt person by person and tool by tool until it evaporates when someone leaves. It sits beneath an organization's applications and above foundational AI models: applications generate knowledge into the layer; humans and AI consume governed context from it.
  3. Term 03The Human LayerThe Human Layer is the architectural layer that connects human judgment with artificial intelligence: the set of structures - decision gates, escalation protocols, accountability structures, override mechanisms and trust calibration interfaces - through which AI systems remain governed by, and accountable to, the humans responsible for their outcomes.
  4. Term 04Context ContinuityContext continuity is the property of a system in which every interaction builds on the accumulated context of previous interactions, rather than starting from zero. In enterprise AI, context continuity means an AI system operates with persistent knowledge of the organization it serves - its history, decisions, relationships and constraints - across sessions, tools and time.
  5. Term 05Decision LineageDecision lineage is the ordered, append-only record of a decision's full life: the recommendation, the action taken, the delta between them, the identity of who decided, and the outcome. It is the atomic unit of organizational memory - distinct from information storage, knowledge management and audit logging, because it preserves not just what was decided but how judgment was exercised.
  6. Term 06Bi-temporal MemoryBi-temporal memory records every fact against two separate timelines: when it was true in the world, known in data modelling as valid time, and when the organization came to know it, known as transaction or knowledge time. Timer calls these its two clocks and applies them to an organization's decisions, commitments and agreements rather than to rows in a database.
  7. Term 07Sovereign Memory LayerThe Sovereign Memory Layer is the architecture through which an organization owns its memory and can prove its judgment, independently of the AI models, platforms and vendors it rents. It binds two assets that are the same thing seen from two sides - the record of what the organization has done and decided, and the accountable human judgment that record makes legible - and keeps both portable across model change, vendor change and personnel change.
  8. Term 08Institutional KnowledgeInstitutional knowledge is the accumulated understanding of how an organization actually works: the reasoning behind its decisions, the exceptions to its documented processes, the history of its relationships, and the judgment calls that experienced people make without being told to. It is held predominantly in individuals rather than in systems, which is why it leaves the organization when they do.
  9. Term 09Institutional MemoryInstitutional memory is an organization's retained record of its own past: what it decided, why it decided it, what happened as a result, and which commitments and constraints still bind it today. It is the organizational equivalent of long-term memory, and in most organizations it is carried informally by long-tenured people rather than held by any system.
  10. Term 10Corporate MemoryCorporate memory is the business term for the knowledge, decisions and context a company retains over time, independently of the individuals who produced them. It is used interchangeably with institutional memory. Corporate memory loss describes the resulting failure state: a company that can no longer reconstruct why it does what it does, typically after turnover, reorganization or a migration between systems.
  11. Term 11Tribal KnowledgeTribal knowledge is an informal industry term for know-how held collectively by a group of experienced employees and never formally written down: the undocumented fixes, tolerances, exceptions and judgment calls that make a process work in practice rather than on paper. In business, manufacturing and software it almost always signals a documentation gap rather than a shortage of expertise.
  12. Term 12Knowledge LossKnowledge loss is the permanent disappearance of an organization's working context: the moment the reasoning behind a decision, the history of a relationship or the cause of a process exception exists only in one person's memory or one retired system, and then becomes unreachable. Its defining property is that it is silent. Nothing fails at the moment of loss; the cost appears later, when the organization tries to reconstruct why.
  13. Term 13Context WindowA context window is the maximum amount of text, measured in tokens, that a language model can consider in a single pass. The instructions, the conversation so far, any retrieved documents and the model's own response all consume it. When the window fills, the earliest content is dropped or summarized, and once a session ends the window is discarded entirely.
  14. Term 14AI Agent MemoryAI agent memory is the set of mechanisms that let an AI agent carry information beyond a single exchange: short-term memory held in the context window, episodic memory of past interactions, semantic memory retrieved from an external store, and procedural memory of how a task is performed. Unlike a model's parameters, which are fixed when training ends, agent memory is written and read at run time.
  15. Term 15AI Audit TrailAn AI audit trail is the append-only record of how an AI-assisted decision was reached: what the system recommended, which context and evidence it drew on, which human reviewed it, what they changed or overrode, and what the outcome was. It differs from a conventional system log in its subject: a log records transactions, while an AI audit trail records judgment.
  16. Term 16AI GovernanceAI governance is the set of structures through which an organization decides how AI may be used, who is accountable for its outputs, and how that accountability is evidenced: policy, risk classification, human oversight, escalation paths, record-keeping and review. It is distinct from AI ethics, which sets the principles, and from AI compliance, which demonstrates conformity with a particular regulation.
  17. Term 17Enterprise AIEnterprise AI is the use of artificial intelligence inside an organization under the constraints that separate enterprises from consumers: shared accountability, regulatory exposure, existing systems of record, data residency requirements, and the expectation that any output can be explained and audited. Its defining difficulty is not model capability but organizational context.

Get started

The future belongs to organizations that remember.

Get started Score your organization

For companies of 50 people or more.