We've come a long way since the TrustGraph site was nothing more than a Docusaurus collection of documentation masquerading as a website! Check out the brand new TrustGraph.ai!

🚀 Join the GOAI Open Source AI Challenge! Calling all AI developers, builders, and open-source enthusiasts! The GOAI Open Source AI Challenge is now open, bringing together developers from around the world to build the next generation of AI applications. 🏆 Global competition with a ~US$730K+ prize pool 🌍 Open to developers worldwide 💻 Build with open-source AI technologies 🤝 Meet fellow builders at the Silicon Valley Roadshow (Aug. 8) Learn more & register: goaihz.com/en Silicon Valley Roadshow: luma.com/cxu2xd95

Who doesn't love breakfast all day?

Watch TrustGraph 2.6 in action: youtu.be/lmhmrJ7zRE0

Have you been wondering, are people actually doing anything with context graphs? Watch our latest YouTube video to not only see context graphs in action, but agents that use them, generate decision traces, and even generate decision analytics. 👇 youtube.com/watch?v=lmhmrJ7z…

TrustGraph 2.5 is out 🎉 The highlights: ☁️ @alibaba_cloud now supported 🔬 SPARQL + GraphQL Workbenches in the UI 🌍 Live workspace/collection/flow switcher in the UI header 🔐 MCP server auth actually works now (Bearer token, in-band WebSocket) ⚡ SPARQL 1.1 engine: 30+ new functions, MINUS operator, streaming eval, bind join optimisation 🗄️ Knowledge cores now preserve full provenance on round-trip 🏭 Cassandra TLS + replication fixes, Qdrant shard/replication config, K8s Secrets for object store creds Bunch of workspace routing and large document OOM fixes

TrustGraph 2.4 is out — and it has a UI now 🎉 New in 2.4: - Full web UI: Agent Console, GraphRAG view, 3D Knowledge Explorer, Ontology Workbench, Schema Workbench, Document Ingestion, Flow Management - JWT-based IAM with API keys, capability-based auth, and full user/workspace management CLI - Workspace multi-tenancy — enforced at the queue layer, not message fields - Pluggable bootstrap framework replacing tg-init-trustgraph - No-auth mode for dev/demo deployments - Async-safe Cassandra + Qdrant I/O - Pulsar message loss + stale producer bugs fixed

How ontologies like W3C PROV-O are the foundation of explainable AI.

TrustGraph 2.3 is live 🎉 What's new: - Processor Groups → up to 2.5 GB memory savings per install - RabbitMQ production-ready → another ~1 GB saved - Kafka backend (beta) - amd64 + arm64 multi-arch containers - LLM token usage exposed end-to-end for cost tracking - Ontology domain/range validation on triple extraction - Agent explainability instrumentation + TrustGraph ontology as Turtle file Up to 3.5 GB lighter. Broader messaging fabric support. Deeper observability.

What are ontologies and how are they relevant to context graphs?