Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: Ontology
Ontology Shape vs Model Size: What a Graph's Design Costs an AI Agent
How three lossless shapes of one insurance knowledge graph changed accuracy, tool calls, tokens and cost for small and large models on Azure Cosmos DB.
30 min readFixing RAG With an Ontology and Knowledge Graph in Semiconductor Docs
Semiconductor EDA docs got an RDF/OWL ontology, a Neo4j knowledge graph and ontology-constrained retrieval, cutting RAG follow-up questions by 30 percent.
8 min readAgents Need Interpretable Meaning: Semantic Layers, Ontologies and Governed Metrics
Conversational BI fails when meaning is unresolved. What text-to-SQL studies show, and how a semantic layer, an ontology and canonical IDs make answers safe.
15 min readOntology or Semantic Layer? What an AI Agent Needs From Both
How ontologies and semantic layers differ, what fails when only one exists, where knowledge graphs fit, and a practical order for building the context layer.
9 min readWhy RAG Isn't Enough: A Semantic Layer for Graph, SQL and Vector Queries
RAG breaks on multi-hop, numeric and ambiguous enterprise questions. A semantic layer handles decomposition, routing and schema mapping, with Apache Ossie for portability.
8 min readRDF to GraphRAG: Practical Ontology Layers with VOO Examples
From IRIs and triples through RDFS, OWL and SHACL: practical ontology layers with VOO examples, when they beat tables and how they steady GraphRAG answers.
9 min readSHACL TDD for GraphRAG Agents: One Executive Cap Rule That Halts Bad Actions
Encode a 30% liability-cap executive-approval policy as SHACL, prove it with pytest, and watch an ontology firewall halt the agent on a $2.3M contract demo.
6 min readOntology-Aware GraphRAG: When Vectors Need Typed Relationships
How identifier anchors, ontology contracts, rank fusion, and cite-or-reject gates fix RAG failures on CVEs, multi-hop ownership, and unspoken graph facts.
12 min readOntology-based context engineering when vector RAG is not enough
When meaning spans systems of record, embeddings alone miss joins. Ontology-based context engineering validates typed facts, provenance, and neighborhood tools for grounded LLM answers.
24 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.