Insights into our knowledge layer.
From the first term to the finished system architecture: our collected knowledge from day-to-day practice structured, searchable and directly applicable.
What just landed on the shelf.
The latest pieces from blog, Library and white papers — sorted chronologically, updated automatically.
)
LibraryWhy Your Data Lake Needs a Semantic Layer
Your data lake probably did what it was built to do i.e., stored the data, lowered storage costs, and gave your data teams a place to land structured, semi-structured, and unstructured information from across the business.
)
LibraryWhy AI Semantic Layers Belong in Your Data Fabric Strategy
Your data fabric has done much of what it was supposed to do. Data moves more easily across systems. Teams can access sources that used to be trapped inside separate platforms, and integration work is no longer as brittle as it used to be.
)
BlogEvaluating Schema-RAG Agents: From Debugging to Business Value
Unlike conventional software with predictable logic, Schema‑RAG Agents are probabilistic systems that require rigorous evaluation to maintain accuracy. Benchmarks anchor their behavior and ensure consistent, business‑ready results.
Find exactly what you're looking for.
Library, blog and white papers — one search field across every format, filtered by format and topic.
Four ways in, one knowledge layer.
Library, blog, glossary, white papers — four reading speeds, one shared subject space. Choose the way in that fits your question.
Our perspective
Concept papers, an ontology primer and Knowledge Engineering 101 — each building on the last.
BlogThe Knowledge Layer
Notes from the engagements — short, honest, technical. Our official blog.
Glossary33 terms
From A to Z: knowledge graphs, semantics, ontologies, Agentic Architecture — precisely defined.
WhitepaperDeeper thinking
Edited PDFs for going deeper — some free, some for an e-mail. With a suggested citation.
From proof to your own hands.
From proof in practice to the platform you drive yourself.
Example engagements
A selection of our engagements at Porsche, CARIAD, STIHL and other pioneers.
Demo
Run d.AP against a real dataset — no setup, straight in the browser.
Semantic Layer vs. Ontology
Semantic Layer, Property Graph, Ontology — the same word, three jobs. The difference at every level, with live proof.
Special insights: digetiers on Air.
Get to know the people behind digetiers and find out what sets us apart. We talk about personal career paths or share our view on the future of management consulting.
)
)
)
)
Terms before data.
AgentAgentic Architecture
A software component that pursues a goal autonomously — through tool calls, with a reasoning trace, tested against an eval suite. In the Knowledge Layer it sits on top of a vocabulary layer.
BFOOntologies
Basic Formal Ontology. Foundational Ontology with top-level categories such as Continuant and Occurrent. Frequently used in industry, pharma and biomedicine.
DeterminismAgentic Architecture
The same input → the same output. LLMs are not deterministic.
DOLCEOntologies
Descriptive Ontology for Linguistic and Cognitive Engineering. Foundational Ontology focused on the cognitive plausibility of its categories.
EmbeddingAgentic Architecture
A vector that represents text mathematically. The foundation of how RAG finds matching answers.
EvalAgentic Architecture
A structured test suite that checks an agent's behaviour against expected statements. Three categories: Functional, Red-Team, Compliance.
Foundational OntologyOntologies
The topmost ontology level, defining generic categories such as object, event, property. It anchors domain-specific models.
GenAIAgentic Architecture
Generative AI. AI systems that generate content on their own: text, images, code or speech.
GFOOntologies
General Formal Ontology. Foundational Ontology from the Onto-Med initiative at Leipzig University.
GlossarySemantics
A controlled list of terms with definitions — the precursor to a formal ontology. The most important negotiation artefact between business domains.
IRIKnowledge Graphs
Internationalized Resource Identifier. A unique identifier for every entity in the graph — the mechanism that ties knowledge together into a whole.
ClassOntologies
A concept type in an ontology. A class describes what two things have in common — not how they happen to be stored.
Knowledge GraphKnowledge Graphs
Fills the ontology with real data — actual customers, contracts, machines. The ontology is the map, the KG is the terrain.
Knowledge LayerKnowledge Graphs
Unites ontology (rules) and Knowledge Graph (facts) into a cross-system single source of truth for meaning. A growing knowledge asset — every new use case extends it.
LLMAgentic Architecture
Large Language Model. A word-guessing machine: it predicts the statistically most likely next word. It formulates language but understands no meaning.
MCPAgentic Architecture
Model Context Protocol. Connects AI agents directly to systems such as SAP or Salesforce. It solves access, but not understanding.
MeshAgentic Architecture
A topology of several specialised agents that communicate through a shared knowledge model and distribute sub-tasks.
Neuro-Symbolic AIAgentic Architecture
LLMs for language + ontology & KG for logic and facts. The LLM understands the question, the Knowledge Layer delivers the fact-based answer — explainable and free of hallucination.
OntologyOntologies
Models concepts, relationships and business rules (e.g. an order must be assigned to exactly one customer). Conceived top-down from the business logic, not from the data model.
OWASemantics
Open World Assumption. Knowledge may be incomplete and can be extended at any time. The basis for the Knowledge Layer as a scalable model.
OWLSemantics
Web Ontology Language. An extension of RDF for more expressive ontologies — including complex rules, classes and relations.
Property GraphKnowledge Graphs
A graph model in which nodes and edges carry arbitrary key-value properties — more pragmatic than RDF, without formal semantics.
RAGAgentic Architecture
Retrieval-Augmented Generation. Gives the LLM access to your own documents as context. Helps with simple questions — structurally out of its depth on complex, cross-system questions.
RDFKnowledge Graphs
Resource Description Framework. Describes knowledge in triples: subject — predicate — object (e.g. customer X — has — order Y). The machine-readable language for ontologies.
Reasoning TraceAgentic Architecture
The traceable path the AI takes to its answer — every statement can be followed back to its source in the Knowledge Graph.
Schema-RAGAgentic Architecture
A RAG variant that uses structures and meanings from the Knowledge Graph instead of text passages.
SemanticsSemantics
The machine-readable description of what data means — not just where it sits. Without semantics, AI reads data but does not understand it.
SHACLSemantics
Shapes Constraint Language. A W3C standard that validates RDF graphs against shape templates. The complement to OWL for structural checks.
SKOSSemantics
Simple Knowledge Organization System. A W3C standard for lightweight concept hierarchies — taxonomies, thesauri, controlled vocabularies.
SPARQLKnowledge Graphs
The query language for RDF graphs. Pattern matching on subject-predicate-object triples, with federated-query support across multiple endpoints.
TripleKnowledge Graphs
A subject-predicate-object statement. The atomic unit of an RDF graph. A statement such as Knowledge Layer has pillar.
Triple StoreKnowledge Graphs
A specialised graph database for RDF triples. Examples: Stardog, GraphDB, Neptune, Virtuoso.
VocabularySemantics
A controlled set of terms for a domain. The precursor to an ontology — it defines what an ontology then formalises.