Unlocking AI-Powered Data Integration with OPAL and Virtuoso

Animesh Kumar's piece on Modern Data 101 names three capabilities that make a knowledge graph AI-ready: ≫ ontological grounding via OWL to kill term ambiguity ≫ subgraph serialization to fit graph slices into context windows ≫ decision-awareness to capture decision traces The graph bridges data products with the LLM ... connecting meaning no single product contains. Knowledge accumulates ... it does not evaporate. The asset no foundation model ships with. #knowledgeGraph #ontology #contextEngineering #dataProducts

LLMs are good at reasoning. They can't know your "net_rev" definition, or that bonus_disbursement is downstream from it. Animesh Kumar on AI-ready knowledge graphs: OWL-typed ontological grounding so the model doesn't guess at domain terms; subgraph serialisation into the context window (SPARQL/Cypher back, RDF Turtle or JSON-LD out); decision-awareness that records why a path was chosen, not just what. LLMs already hold the personalisation layer. Supply the semantics. #knowledgeGraphs #ontology #semanticLayer #LLM #dataProducts

Francisco Marques da Silva frames what most agentic workflows silently pay: a knowledge retrieval tax. Every session re-explains domain constraints from scratch, producing plausible output that drifts across builds. His fix: version domain knowledge as queryable artifacts … taxonomies, schemas, ontologies … that agents consume via API instead of re-inferring from chat. Built once, reused everywhere. Consistency emerges from structure, not memory. #metadataDrivenDev #ontology #agenticAI #contextEngineering

AI Finding Your Data Taken from the DM Radio Podcast from Feb 27, 2026 💡Ask Questions, Get Answers - No SQL Required Live demo of Ask Collate finding customer data intelligently based on meaning, not just name matching With Suresh Srinivas, Sriharsha Chintalapani, and @pmbrull #knowledgegraph #ontology #semantics #openmetadata #askcollate

Models drift.  Context gets lost between steps.   Logic has to be rebuilt in every layer. We’ve been thinking about it differently with the EvolveAI Suite. EvolveOntology, EvolveBuilder, EvolveTrainer, and EvolveAgents operate on the same shared model —   so context, logic, and meaning carry through by default. No remapping. No translation gaps. More to come. evolveai.tools #EvolveAI #EnterpriseArchitecture #AgenticAI #AIi #Ontology #KnowledgeGraph

Another release, another round of "why didn't this exist from the start?" ORIONBELT Ontology Builder v1.2.0 (OBOB :D) now ships with built-in upper ontology support. First up: gist by Semantic Arts. Pick your modules, hit load, done. No more copy-pasting Event, Person, and Organization classes for the 47th time. Because honestly, if you're still hand-crafting foundational concepts from scratch in 2026, we need to talk. Also in this release: consistent tab navigation that actually remembers where you clicked, branded UI with our logo blue, and View/Edit buttons that stop playing hide-and-seek. Try it: https://lnkd.in/dERFebCy GitHub: https://lnkd.in/egGZJHiP #ontology #semanticweb #knowledgegraph #owlontology #linkeddata #gist #opensource #obob

a few years ago I was nerding out on knowledge graphs, which have become much more popular as a way to make AI systems perform better. Thought I'd resurface this interview with Amy Hodler - if you ever get to see her speak she's a rockstar, as this episode shows. Worth a listen if you're curious about knowledge graphs and ontologies. #AI #knowledgegraph #ontology #graphdatabase #artificialintelligence Colin Whitlatch Taylor Last Dennis DiPalma Julian Berman Joshua Kohn Jay Snyder Quantum Rise https://lnkd.in/gXNHCBNN

Oxigraph is an RDF graph database library written in Rust. Its goal is to provide a compliant, safe and fast on-disk graph database. It also provides a set of utility functions for reading, writing, and processing RDF files. https://lnkd.in/eqAiZh_u #RDF #Ontology #GraphDatabase

Are we turning ontology into technological scholasticism? In medieval scholastic debates, philosophers famously argued about how many angels could dance on the tip of a needle. Today, something similar is happening in tech. The discussion around “what counts as an ontology,” how many types there are, and how they should be classified has become increasingly abstract—often detached from real engineering needs. Endless definitions, sub‑definitions, meta‑levels, and taxonomies of taxonomies. At some point, it starts to feel like a modern version of the same question: How many kinds of ontologies can fit on the tip of a taxonomy? When conceptual debates stop improving our tools and start orbiting themselves, we’ve crossed from useful theory into technological scholasticism. Maybe it’s time to shift the focus back to what ontologies are meant to do: help us build systems that understand, reason, and act. #ontology #semanticontology #knowledgeengineering #semanticweb #knowledgegraphs #rdf #owl #linkeddata #ontologies

Flexing my agent building skills at CATIA Open House 2026 😁 Showcased how three AI agents can simplify everyday engineering workflows - from understanding a new feature to delivering and debugging it using real project data. The focus was on making workflows faster, more intuitive and less dependent on manual digging. #AIFirst #Productivity #CATIA #AIAgents #Ontology #Deterministic #LLMAgnostic