Building GraphRAG Agents with ADK | Google Codelabs
This codelab teaches you how to build intelligent multi-agent systems using Google’s Agent Development Kit (ADK) combined with Neo4j graph databases and the Model Context Protocol (MCP) Toolbox. You’ll learn to create specialized GraphRAG-powered agents that leverage knowledge graphs for context-aware query responses, implement agent orchestration patterns, and deploy pre-validated database queries as reusable tools. By the end, you’ll have built a production-ready investment research system demonstrating best practices for next-generation retrieval agents.
A research project on the history of knowledge graphs using Gemini's Deep Research Agent
I conducted a deep research project on the history of knowledge graphs using Gemini's Deep Research Agent and extracted a knowledge graph from the research report in one pipeline.
You can find the code that I used to build the agent inside a Colab notebook labelled "KG_Research_Agent.ipynb" inside this GitHub repository:
https://lnkd.in/dXsYb_4V
I gave the agent the simple prompt: "Research the history of knowledge graphs", and it came up with a detailed report, along with references, which you can read here: https://lnkd.in/erv-aPqK
Connected Data Oxford Semantic Technologies Google DeepMind Gephi GitHub
#knowledgegraphs #agents #deepresearch #semanticweb
esearch project on the history of knowledge graphs using Gemini's Deep Research Agent and extracted a knowledge graph from the research report in one pipeline.
Learn graph database fundamentals, query languages, and implementation patterns for AI systems. Covers GraphRAG, knowledge graphs, and performance optimization.
Theoretical foundations and engineering tools for building ontologies as reference conceptual models
PDF | Perhaps the most fundamental notion underlying the desiderata for a successful Semantic Web is Semantic Interoperability. In this context,... | Find, read and cite all the research you need on ResearchGate
A Response to Critical Discourse on Ontology-Grounded Knowledge Graphs Yesterday I published an article on the engineering realities of building BFO-grounded knowledge graphs at scale. The response from the enterprise architecture community surfaced a critical architectural question that deserves de
Engineering Reality: What It Actually Takes to Build a BFO-Grounded Knowledge Graph at Scale | LinkedIn
Beyond the Hype: A Practitioner's Perspective on Ontology-First Architecture The knowledge graph market has exploded with vendors claiming semantic capabilities, ontological reasoning, and enterprise-grade inference. Yet a significant gap exists between marketing collateral and engineering reality.
Context Graphs: What the Ontology Debate Gets Wrong | LinkedIn
The context graph conversation has exploded over the past two weeks. Foundation Capital's "Context Graphs: AI's Trillion-Dollar Opportunity" kicked it off.
Generating Taxonomies Promptly: Practical LLM applications for human-centric taxonomy development | LinkedIn
Co-authored with Gigi Shannon Every taxonomist knows the pressure of delivering the most accurate models under tight timelines and tighter budgets. Now that AI has permeated enterprise tools and strategic plans, the question isn’t whether they should use LLMs, but how to make them work most effectiv
released the GraphStorm v0.5.1!This is a milestone. Now GraphStorm is nearly ready for most Graph Machine Learning tasks, including real-time inference that is critical for time-sensitive tasks, e.g., fraud prevention.
Ontologies, Business Data Models, and Where Semantics Actually Break
Ontologies, Business Data Models, and Where Semantics Actually Break
This post by Pierre Bonnet raises real concerns that many enterprises are feeling right now. Knowledge graphs and AI have made semantic modeling accessible, but not always disciplined. Poorly built ontologies absolutely can create silos rather than remove them.
That said, there is an important distinction worth clarifying.
A Business Data Model is a critical enterprise artifact. It captures what an organization chooses to track, report on, and operationalize. It is purpose driven, scoped, and optimized for business outcomes.
An ontology, properly used, serves a different role. It is not a replacement for business data modeling, nor is it a technical derivative of it. Ontology constrains meaning at a deeper level by defining what kinds of things exist in a domain, how they persist, and how they relate independent of systems, workflows, or reporting needs.
Several of the “mistakes” identified here are exactly why ontology exists in the first place:
• Modeling workflows instead of stable domain entities leads to brittle systems
• Collapsing conceptual meaning into logical or technical artifacts obscures intent
• Treating metadata as documentation rather than as part of a formal semantic structure limits reuse and reasoning
Where the framing becomes risky is in suggesting that ontology should be derived from business data models. In practice, that reverses the dependency.
When business models are built without ontological constraints, today’s organizational assumptions get hard coded into tomorrow’s AI systems. That is where semantic debt and long term integration failures come from.
The most durable pattern looks like this:
• Ontology constrains meaning and category boundaries
• Domain and business models specialize those constraints for enterprise use
• Logical and physical models optimize for execution and performance
When this order is respected, ontologies reduce silos rather than create them, and AI systems gain stability rather than fragility.
The real issue is not ontology versus business data modeling. It is whether meaning is treated as foundational infrastructure, or as a byproduct of implementation.
That distinction determines whether semantic systems scale beyond pilots and remain profitable over time.
#Ontology #SemanticInfrastructure #EnterpriseArchitecture #KnowledgeGraphs #DataStrategy #AIGovernance #SemanticAI #DataModeling #ExplainableAI #DigitalTransformation | 18 comments on LinkedIn
Ontologies, Business Data Models, and Where Semantics Actually Break
The Missing Layer in Enterprise Ontologies Enterprise software got very good at storing state, but it’s still bad at storing decisions. Most systems can tell you what’s true right now and what happened, but they don’t preserve why a choice was made in the moment—what inputs were considered, which co
Graph Retrieval-Augmented Generation: A Survey | ACM Transactions on Information Systems
● First comprehensive GraphRAG overview. This paper provides the first systematic survey of Graph‑based Retrieval‑Augmented Generation (GraphRAG) -- where structured graph data (nodes/relationships) is used to enhance LLM outputs by enabling richer, multi‑hop reasoning compared with text‑only RAG.
● Formalised workflow across three stages. It defines a GraphRAG pipeline -- (1) graph‑based indexing, (2) graph‑guided retrieval (using non‑parametric, LLM or graph neural retrievers) and (3) graph‑enhanced generation -- improving precision and relational context in LLM responses.
● Practical impact across domains. The survey highlights how GraphRAG boosts complex tasks -- from question answering and recommendation systems to domain‑specific reasoning in healthcare, finance and e‑commerce -- by bridging graph knowledge and large model reasoning.
An OWL ontology and RDF knowledge graph of the top 100 IMDb movies, modeled in Protégé and stored in GraphDB. Demonstrates semantic modeling, data ingestion, and SPARQL querying.
An OWL ontology and RDF knowledge graph of the top 100 IMDb movies, modeled in Protégé and stored in GraphDB. Demonstrates semantic modeling, data ingestion, and SPARQL querying. - marcusv02/film-k...
