Knowledge Management Systems — Foundations

Why this is in the vault. The vault, the portal, the team alignment process — all of it is a KMS, intentionally designed. Understanding the field properly clarifies why we do what we do, what’s load-bearing vs decoration, and where the gaps are. Anchor doc for the Knowledge Management/Concepts/ folder.


1. What KMS is

A Knowledge Management System combines people · process · technology to deliberately create, capture, organize, share, and apply knowledge across an organization — so the organization gets faster, smarter, and less dependent on any single person.

Not a document repository. Not a search engine over files. Not a wiki. Those are tools that support KMS. The discipline itself is about turning individual know-how into organizational capability.


2. Five foundational concepts

Each has a dedicated note in ./Concepts/:

#ConceptNote
1The hierarchy of data → wisdomDIKW Pyramid
2What knowledge is (Polanyi 1966)Tacit vs Explicit Knowledge
3How knowledge converts (Nonaka 1995)SECI Model
4The operational loopKM Lifecycle
5The personal–organizational splitPersonal Knowledge Management

Read those five before anything else if KM is new to you. ~20 minutes total.


3. Classification of KMS (Alavi & Leidner 2001)

The canonical academic taxonomy. See KMS Classification.

TypePurposeNOTQIN today
Knowledge RepositoriesStore explicit knowledge✅ The vault
Knowledge Directories / Expert LocatorsTell you who knows what⚠️ People Map (underused)
Knowledge Networks / CoPsConnect people for socialization🟡 Team Discord (TBC at kick-off) — see Communities of Practice
Knowledge Application SystemsEmbed knowledge into workflows⏳ Future Smartex Agent (NOTQIN Agent)

4. Adjacent / overlapping disciplines

KMS borrows from and feeds into many fields. Map of which sibling discipline contributes what:

DisciplineContribution to KMS
Library / Information ScienceCataloging, taxonomy, controlled vocabularies, faceted classification
Information ArchitectureSite structure, navigation, findability
Taxonomy & Ontology EngineeringHierarchies + relationships (RDF, OWL, schema.org)
Knowledge GraphsEntities + relationships + reasoning — the 3D version of UNS
Semantic WebMachine-readable knowledge (RDF, SPARQL, linked data)
Enterprise Content Management (ECM)Document lifecycle, retention, compliance
Learning Management Systems (LMS)Structured training delivery
Business Intelligence (BI)Information → insight via dashboards
Decision Support Systems (DSS)Knowledge → decision in context
Organizational LearningLearning Organization — why teams that learn outcompete
After-Action ReviewsUS Army origin — structured learning from any event
Information Retrieval (IR)Search algorithms, ranking, relevance
Augmented IntelligenceLLMs as retrieval prosthesis — see Augmented Intelligence and RAG
Innovation ManagementKM as input to creating new value

5. The modern technology stack

What goes into a serious 2026 KMS:

LayerPurposeExamples
CaptureGet knowledge inNotes apps, voice→text, meeting transcribers, screenshots
StoragePersist itMarkdown + git, document DBs, vector DBs
Linking / structuringCreate relationshipsWikilinks, Knowledge Graphs, ontologies
RetrievalFind it laterFull-text + semantic search, RAG
SynthesisCombine piecesMaps of content, AI summarization
DistributionRight person sees itPortals (Quartz), email digests, audience routing
Access controlWrong person doesn’tRBAC, Cloudflare Access, audience tags
VersioningTrack changesGit, dated entries
Workflow integrationKnowledge in the tools of workIDE plugins, Slack integrations, MES embeds

6. The AI-era shift (2023–2026)

Three changes that broke / re-shaped classical KM:

  1. RAG — LLMs as a query layer over your knowledge base. Reading the vault through Claude Code IS RAG-by-conversation.
  2. Tools-of-thought — Obsidian, Roam, Logseq, Notion AI, Mem. PKM became dramatically more powerful. Org KM is still catching up.
  3. Agent memory — MCP + persistent memory blur the line between KMS and the tool that uses it. The MEMORY.md file is an example.

Strategic implication for NOTQIN: tacit-to-explicit conversion got cheaper (transcription + LLM summarization), and synthesis (Combination, see SECI Model) is now near-free. The high-value investment is capturing tacit knowledge from operators and customers because everything else is now commodity.


7. How NOTQIN’s vault maps to KMS theory

Vault constructKM concept
Markdown files + gitKnowledge Repository
audience: frontmatter + Quartz portalDistribution layer with RBAC
Wikilinks + backlinksKnowledge graph (implicit)
MEMORY.md (Claude auto-memory)AI agent memory / RAG context
NOTQIN Industrial Intelligence Operating Thesis 2026Externalization of the founder’s mental model
Industrial Intelligence/Concepts/* (OEE, MTBF…)Atomic explicit-knowledge units (Zettelkasten-style)
Discovery Calls/<co>-<date>.md (when populated)Externalization of customer interactions
Wave-end retros + Friday demosAfter-Action Reviews / lessons-learned ritual
Weekly review templateRefine phase of the KM Lifecycle
Sector Pack Audits/Knowledge embedded into code artifacts
Smartex Agent (future)Knowledge Application System
The team itself (Mon/Wed/Fri rituals)Communities of Practice
People MapKnowledge Directory (expert locator)
Onboarding PackInternalization scaffold (see SECI Model)
Decision LogOrganizational memory anchor

The honest read: the vault runs ~80% of a serious org KMS. The gaps are tacit capture from customers/operators (Discovery Calls folder still empty), Expert Locator under-investment, and Knowledge Application (Smartex Agent not yet decision-embedding).


8. Reading list, ranked by NOTQIN leverage

#WorkAuthorWhy for you
1The Knowledge-Creating CompanyNonaka & Takeuchi (1995)The SECI Model source. Foundational.
2Working KnowledgeDavenport & Prusak (1998)The practitioner’s bible. Tacit knowledge in orgs.
3Building a Second BrainTiago Forte (2022)Personal Knowledge Management bible. CODE method.
4How to Take Smart NotesSönke Ahrens (2017)Zettelkasten Method for atomic-note KM. Maps to your concept notes.
5The Fifth DisciplinePeter Senge (1990)Learning Organization foundations.
6Cultivating Communities of PracticeWenger / McDermott / Snyder (2002)Communities of Practice canonical reference.
7Information Architecture for the World Wide Web (4th ed.)Rosenfeld & MorvilleThe IA bible. Your folder structure IS Information Architecture.
8Knowledge Management Systems: A Perspective on the Field (MIS Quarterly 2001)Alavi & LeidnerAcademic anchor. Free PDF. PFE-defensible citation. See KMS Classification.
9Personal KnowledgeMichael Polanyi (1958)Origin of Tacit vs Explicit Knowledge.

For the PFE viva: anchor #1, #2, #8 by name — these are the citations Moroccan examiners recognize.


9. Anti-patterns to avoid

Anti-patternWhy it kills KM
Hoarding without applyingKnowledge that is never re-used is decoration, not capability
Bureaucracy without trustIf people fear capture, tacit knowledge stays hidden
Combination-only loopEndless synthesis from existing docs without new external/internalization → the system inbreeds
One-way pipeline (capture → forget)Without the Apply phase the KM Lifecycle is broken
Tool-first thinking”Let’s adopt Notion” is not a KM strategy. The discipline comes first.
No access control disciplineEither too open (sensitive leaks) or too closed (capability hoarded)
Confusing documents with knowledgeA 200-page Word doc nobody reads is not “captured knowledge”
No expert locatorIf “ask the person who knows” requires asking who that person is, KM has failed at Directory

See also