# singh-kdlc-knowledge-development-life-cycle-2026-06-28

## Veille

Third installment of Ashish Singh's « New Engineering Disciplines for the AI Era » series, devoted to **KDLC — Knowledge Development Life Cycle**: an **8-stage** life cycle for turning enterprise knowledge into an **engineered asset**, on a par with code or data. Thesis: AI initiatives fail not for lack of the right LLM choice or a deployed RAG system, but because they **do not address the underlying structure of knowledge** — « AI is only as effective as the knowledge it can discover, understand, retrieve, and trust ». The KDLC chains Discovery → Extraction → Structuring → Knowledge Graph → Embedding → Index Optimization → Retrieval Evaluation → Refresh. It contrasts **traditional RAG** (isolated documents, keywords) with the **Enterprise Knowledge Fabric** (Knowledge Graphs + Semantic Search + Vector DB + Hybrid Search), where agents understand « relationships, context, and business meaning ». Signature line: « Models provide reasoning. Memory provides continuity. Knowledge provides understanding. » Three examples (finance/compliance, software engineering, healthcare) illustrate the impact.

## Titre Article

New Engineering Disciplines for the AI Era Part 3: KDLC — Knowledge Development Life Cycle

## Date

2026-06-28

## URL

https://www.linkedin.com/pulse/new-engineering-disciplines-ai-era-part-3-kdlc-knowledge-ashish-singh-3sjxc/

## Keywords

KDLC, knowledge development life cycle, knowledge life cycle, enterprise knowledge fabric, knowledge engineering, knowledge engineering, RAG, retrieval-augmented generation, knowledge graph, knowledge graph, semantic search, semantic search, vector database, vector database, hybrid search, embedding, index optimization, retrieval evaluation, knowledge refresh, agentic AI, AI agents, engineered asset, enterprise data, knowledge governance

## Authors

Ashish Singh

## Ton

Profile: a technical-leadership LinkedIn article (third installment of a numbered series, « New Engineering Disciplines for the AI Era »), an architect/consultant perspective in the third person, an analytical and prescriptive register aiming to be a reference framework. Target audience: data/AI leaders, enterprise architects, platform decision-makers confronted with the recurring failure of AI projects. The tone is that of a **new discipline being instituted**: it does not describe a tool but elevates "knowledge engineering" to the rank of a full engineering discipline, by explicit analogy with the SDLC and the data life cycle. The rhetoric proceeds by **reversing the diagnosis** (the problem is not the model but the knowledge) then **structuring it into a numbered cycle** (8 stages), a classic LinkedIn-framework device — each stage is named, defined, ordered, which gives an impression of methodological completeness. Memorable, triadic phrasing (« discover, understand, retrieve, and trust »; « Models provide reasoning. Memory provides continuity. Knowledge provides understanding. ») carries the authority more than figures do — the article is **conceptual, non-quantified**, with no benchmark or cited study. The opposition between "naive" RAG and the Enterprise Knowledge Fabric serves as the argumentative pivot: RAG is the simplifying foil the discipline moves beyond. Sector grounding (finance, engineering, healthcare) is illustrative rather than demonstrative in register.

## Pense-betes

- **Reversal thesis**: AI projects fail not from a poor LLM or RAG choice, but from a **lack of structured knowledge**. « AI is only as effective as the knowledge it can discover, understand, retrieve, and trust. » → the lever sits upstream of the model.
- **KDLC = 8 stages** (worth remembering in order): 1. **Knowledge Discovery** — locate knowledge scattered across data (DBs, SharePoint, wikis, CRM, ERP, engineering artifacts, comms channels). 2. **Knowledge Extraction** — extract while **preserving business context, metadata, relationships, ownership**. 3. **Knowledge Structuring** — standardized, reusable knowledge objects. 4. **Knowledge Graph Creation** — map customers ↔ products ↔ projects ↔ teams ↔ regulations ↔ applications. 5. **Embedding** — semantic representations (understanding by meaning, not keywords). 6. **Index Optimization** — vector indexes, metadata, retrieval pipelines. 7. **Retrieval Evaluation** — measure relevance, precision, completeness, **business impact**. 8. **Knowledge Refresh** — continuous updating (policies, regulations, releases).
- **RAG vs Enterprise Knowledge Fabric**: traditional RAG retrieves **isolated documents** by keyword; the **Knowledge Fabric** combines Knowledge Graphs + Semantic Search + Vector DB + Hybrid Search → agents understand **relationships, context, and business meaning**, not just documents. RAG = foil, Fabric = target.
- **Founding triad**: « **Models provide reasoning. Memory provides continuity. Knowledge provides understanding.** » → knowledge is the "understanding" layer, distinct from reasoning (model) and continuity (memory).
- **Disciplinary positioning**: knowledge engineering becomes **as critical as software engineering and data engineering** in the age of agentic AI.
- **3 sector examples**: financial compliance (linking regulations ↔ internal policies), engineering assistant (architecture + API contracts + standards + incidents before recommending), clinical assistant (guidelines + protocols + literature + patient record).
- **Limitations**: a **conceptual, non-quantified** article (no benchmark, cost figures, or measured field feedback); the KDLC is an appealing framework but remains a **taxonomy of stages** whose implementation, governance, and maintenance cost (stage 8, Refresh) are the real pain points, left untreated.
- **Related**: a direct parallel with **SDLC/ADLC** (Hingel six-stage, Williams ADLC, SFEIR 11 phases) — the same « life cycle » move applied to knowledge instead of code; **RAG in structural decline** (KB entity) that the Knowledge Fabric claims to move beyond; converges with the **Compounding Knowledge Lifecycle** pattern and the logic of the knowledge graph (this repository's ontologie-kg is a concrete instance of it).

## RésuméDe400mots

Third installment in the « New Engineering Disciplines for the AI Era » series, this article by Ashish Singh establishes **KDLC — Knowledge Development Life Cycle** as an engineering discipline in its own right. Its thesis reverses the dominant diagnosis: if so many enterprise AI initiatives fail, it is not for lack of choosing the right model or deploying a RAG system, but because they ignore the **underlying structure of knowledge**. The pivotal line sums up the stakes: « AI is only as effective as the knowledge it can discover, understand, retrieve, and trust. » Knowledge must therefore be treated as an **engineered asset**, on a par with code (SDLC) or data.

The KDLC organizes this work into **eight** ordered stages. **Discovery** locates knowledge scattered across databases, SharePoint, wikis, CRM, ERP, and engineering artifacts. **Extraction** pulls out the meaningful information while preserving business context, metadata, relationships, and ownership. **Structuring** converts the informal into standardized, reusable knowledge objects. **Knowledge Graph Creation** maps the interconnections among customers, products, projects, teams, regulations, and applications. **Embedding** produces semantic representations enabling understanding by meaning. **Index Optimization** refines vector indexes and retrieval pipelines. **Retrieval Evaluation** measures relevance, precision, completeness, and business impact. Finally, **Knowledge Refresh** keeps the whole up to date against new policies, regulations, and releases.

The argumentative core contrasts two architectures. **Traditional RAG** retrieves isolated documents via keyword-based searches. The **Enterprise Knowledge Fabric** — a combination of Knowledge Graphs, Semantic Search, Vector Databases, and Hybrid Search — aims for interconnected understanding: rather than retrieving documents, AI agents grasp « relationships, context, and business meaning ». A second triad frames the respective roles of the layers: « Models provide reasoning. Memory provides continuity. Knowledge provides understanding. »

Three sector illustrations make the impact concrete: a **financial compliance** assistant linking up-to-date regulations and internal policies; a **software engineering** assistant consulting architecture, API contracts, standards, and incidents before recommending; a **clinical** assistant cross-referencing treatment guidelines, protocols, literature, and patient records. Singh concludes that, in the age of agentic AI, knowledge engineering becomes as critical as software engineering and data engineering. The article's limitation lies in its **conceptual, non-quantified** nature: no benchmark or cost figures, and the real pain points — governance and ongoing maintenance of the Refresh stage — remain out of scope.

## GrapheDeConnaissance

- Ashish Singh —a_créé→ New Engineering Disciplines for the AI Era Part 3: KDLC (DOCUMENT, 0.96)
- KDLC —s_applique_à→ connaissance d'entreprise comme actif ingénieré (CONCEPT, 0.93)
- Ashish Singh —affirme_que→ les initiatives IA échouent faute de connaissance structurée, pas faute de modèle ou de RAG (AFFIRMATION, 0.92)
- Ashish Singh —affirme_que→ « AI is only as effective as the knowledge it can discover, understand, retrieve, and trust » (CITATION, 0.94)
- KDLC —fait_partie_de→ New Engineering Disciplines for the AI Era (série) (CONCEPT, 0.9)
- Enterprise Knowledge Fabric —surpasse→ RAG (TECHNOLOGIE, 0.88)
- Enterprise Knowledge Fabric —utilise→ knowledge graph (TECHNOLOGIE, 0.9)
- Enterprise Knowledge Fabric —utilise→ recherche sémantique (TECHNOLOGIE, 0.88)
- Enterprise Knowledge Fabric —utilise→ base de données vectorielle (TECHNOLOGIE, 0.88)
- knowledge graph —permet→ compréhension des relations, du contexte et du sens métier (CONCEPT, 0.87)
- KDLC —s_applique_à→ IA agentique (TECHNOLOGIE, 0.86)
- ingénierie de la connaissance —est_instance_de→ discipline d'ingénierie critique au même titre que SDLC et data engineering (CONCEPT, 0.85)
- Ashish Singh —affirme_que→ « Models provide reasoning. Memory provides continuity. Knowledge provides understanding. » (CITATION, 0.9)
- KDLC —observé_dans→ conformité financière, ingénierie logicielle et santé (CONCEPT, 0.83)

---
Canonical: https://www.thekb.eu/en/fiches/singh-kdlc-knowledge-development-life-cycle-2026-06-28/