I never expected to see the @TrustGraphAI logo next to Minecraft. 😂 Thanks for the support @Olares_OS!

Everyone is talking about ontologies. Why now? What changed? The economics. Ontologies are far from new. If we're talking about semantic web standards like RDF, they date back to the late 90s. OWL, the standard for ontologies, goes back to 2004. What's really changed is how you develop ontologies. The old way: - Read lots of information and try to understand what it all means, mapping out relationships and terms - Try to reach consensus with others on the mappings, or in other words, argue a lot about it - Build all that into a machine-readable structure like OWL As you can imagine, this is a long process, for a complex domain likely months. So what's the new way? - LLMs can read an interpret the domain - LLMs can build the relationships and mappings - LLMs can structure all that in OWL, or any target format you'd like And how long does that take? Minutes. I know what you're probably thinking, what about the quality? Surely human developed ontologies are better, right? Not really. Since we released ontology features and OntologyRAG in @TrustGraphAI, we've studied a lot of ontologies and developed a lot of our own. LLM developed ontologies? Easily on par with human developed ones, and more often than not, better. But where do we go from here? Expect to see even more advanced capabilities in TrustGraph very soon that will be able to develop custom ontologies by learning your business logic through your own data with all the lifecycle change management capabilities you need. Don't trap your business logic in proprietary ontologies. Develop you own with TrustGraph.

I'm excited to be a part of the @DevNetwork_ Advisory Board and be a panelist on Sept 3 of one of the largest gatherings of API, cloud, AI, and enterprise technology leaders driving the next generation of software innovation. Next week three major events come together: ☁️ API World + CloudX + AI TechWorld (September 1–3 | Santa Clara Convention Center) As part of the Advisory Board, I have a limited number of free and discounted passes available for my network: 👉 Free OPEN Pass: link.devnetwork.com/MLtDIly6 👉 25% Off PRO & PREMIUM Pass Code: Advisory-25 If you're building with APIs, AI, cloud infrastructure, agentic systems, or enterprise software, this is a great opportunity to connect with developers, engineering leaders, architects, founders, and technology decision-makers. Let me know if you'll be attending—I'd love to connect in person to discuss how @TrustGraphAI can transform your agentic roadmap! 🗓️ Check out the schedule: cloudxconf.com/conference/

If we really want to talk about graph engineering, shouldn't we be talking about semantic significance?

What Is Graph Engineering? (Hint: It's Not What You've Been Told)

The path to #EnterpriseAI follows a road paved by safety engineering, human factors, cyber risk quantification, behavioral economics, and prospect theory. Together these methods can generate measurable determinism for agentic AI. All coming very soon in @TrustGraphAI...

Today, I read a LLM license so you don't have to! Today's model? Kimi K3. What's the TLDR? Yes, Kimi K3 is open weights, but it's not true open source as it comes with commercial restrictions. It's not MIT, Apache 2, AGPL-3, or even SSPL. It's their own custom written license. It sorta reminds me of some of Meta's licenses for Llama with one big exception... I actually don't understand this license - kinda at all - and I wonder if that's the intent. I also wonder if they had US counsel to help them with the wording, because lost in translation could also be an issue. The key points: - If you provide Model as a Service to a third party with Kimi K3 *and* exceed $20M USD (or equivalent) in revenue over 12 consequetive months, you must obtain a separate - and I'm assumed paid - license from Moonshot AI. - However, there are exceptions to Model as a Service, specifically "...(a) end-user products with model capabilities solely embedded within specific features or harnesses, or (b) mere relaying of requests o models hosted by others." I very much don't understand exception (a), and as Mark Adams and I wondered, if we deploy Kimi K3 within @TrustGraphAI (the deterministic context harness for agents), are we exempt? I think we can argue yes, but I have a feeling that wasn't their intent? - Much like Meta's old licenses for Llama, if you exceed 100 million MAU (montly active users) or $20M USD revenue threshold, "Kimi K3" must be prominently in the UI. - Then there's point 4. Point 4 states that neither the Model as a Service nor MAU thresholds apply if "...(a) internal use of the Software, defined as any use that does not make the Software, its outputs, or its underlying capabilities available to third parties..." This again, is very ambiguous and open to interpretation in my view. If an enterprise were to take a platform like TrustGraph, and use it to deploy Kimi K3, it seems they would be exempt from the Model as a Service and MAU requirements, even if someone (like us) provided engineering services to them to manage/host the TrustGraph deployment that forms the harness for the internal usage. Again, ambiguous. It seems this license is aimed at someone that deploys the model, then provides simple API access to it. However, if you deploy it within a more sophisticated context harness, it's exempt from the revenue threshold? For me, the interpretation is unclear, and that could be the intent. Perhaps they have intentionally released Kimi K3 with an ambiguous license to scare people away from trying to commercialize it. Is Kimi K3 an open weights model? Yes. Is Kimi K3 true open source software? No. There are commercial restrictions that limit usage. Is this the future direction of open weights models? Restrictive licenses that allow us to see the weights but prevent commercialization? We shall see.

Open weight models and open source AI seem to be on everyone's mind. It was a central theme to AMD's Advancing AI conference with Lisa Su's keynote talking about the need to cultivate open ecosystems. We at @TrustGraphAI give our strongest support in advancing open source AI.

What does the World Cup have to do with building in AI? Congratulations to Spain for proving they are the best team, and that a true team effort is what wins championships. While teams like Argentina, France, and England are built around star players like Mbappe, Messi, and Bellingham, Spain doesn't focus on superstars. They focus on the team. A team of players that know their roles. A team of players all on the same page. A team of players working together as a cohesive unit with one goal in mind: team victory. Sure, the superstar players make for exciting moments and great content, but they all went home without what they really wanted - the World Cup championship. And this is how people are approaching AI, treating the LLM like it's the superstar player. Sure, you can build around a LLM as if it's your star player, but will you have an end product that will be usable by large enterprises. No. Enterprise-grade AI products require a team effort. The LLM is just one of many players. In fact, the smaller role you give the LLM, the better your chance is at success. The best AI products are not built to feature LLMs. The best AI products are built to minimize LLMs. But how do you do that? Context engineering. But how do you do that with closed models from Anthropic, OpenAI, or Google that try and lock you into their team's way of doing things? Open Source AI. Not too long ago, the thought of building complete enterprise grade products with open source AI seemed fantasy. But it is reality today. How do I know? Because we do it every day with @TrustGraphAI. TrustGraph is the complete context engineering platform for open source AI. From powering agents with holonic context graphs, to deploying open weight models, to modular and reusable units of context called context cores, to true enterprise-grade multi-tenancy and fine-grained access controls for users, agents, and data, the future of open source AI is now. Not only is TrustGraph built for open source AI, it's open source too. Try it today 👇

Have you heard the term "explainable AI", but wonder what does it actually mean? This is explainable and open source AI 👇

Agentic commerce is much more than just shopping assistants. Retailers and brands are constantly searching for they why of purchases, why with all the available choices, did a consumer make those choices? What options did they see? What options were presented first? How do they make decisions in comparisons to different product options? Do they even care about their stated budget? You can see how all of this works - a lot more - in the @TrustGraphAI Playground Preview, available now.

So much talk lately from Palantir and Nvidia about open models and owning your data... And we agree. Which is pretty odd for us to agree with Palantir on anything, considering we've designed TrustGraph to be the AI-native, anti-Palantir. As one Palantir user told us, "Palantir was not designed for use with AI..." So what makes TrustGraph so different from Palantir? Let's count the ways: 1. TrustGraph is totally open source. The last time I tried to access Palantir's "open" platform, it required multiple forms of government ID to get access. 2. TrustGraph is built on source tech and standards like RDF, OWL, SHACL, etc. 3. TrustGraph has full observability and telemetry to give you real-time insights into complete system performance. 4. TrustGraph has a full LLM inference stack that runs on Nvidia, AMD, and even Intel hardware. Not only can you maintain total control over your data, you can own the model deploys on your own hardware or any environment of your choosing. There are more, but I thought, do I actually need to go on at this point? Instead, give TrustGraph a try today. Open source. Open standards. Total Transparency. Oh, while you're on our GitHub, don't forget to sign up for our Playground Preview, where you will be able to test all the capabilities of TrustGraph directly through your browser, no code, nothing to deploy, no API keys required.

What happens when you have granular explainability tools in your AI stack? You realize how bad LLMs still are. With the realize of version 2 of @TrustGraphAI, we introduced comprehensive explainability tools utilizing semantic web standards like RDF, OWL, SHACL (coming soon), with reification from RDF 1.2. These tools allow fine-grained traceability for agentic workflows. And what do they show us? LLMs have not advanced nearly as much as people claim. What are some common errors: - LLMs are still really bad at semantic filtering. When asked to rank and prioritize context, they often ignore the correct information for seemingly unrelated information, despite the correct information being provably available. - What works with one LLM won't work with another. An interesting use case for fine-grained explainability is LLM comparison testing. Take the same workflow, and the same data, and marvel at how the LLMs work through the data in totally different ways. - LLMs are still really inconsistent. Even when provided with the same context, LLMs will make different decisions using that data, leading to different results. Think you can prompt your way out of these problems? Maybe, but with TrustGraph you have the tools to do exactly. Full traceability and audibility of every agentic action. A/B test your prompts, agent design, agentic patterns, and even how graph context is queried. If you need deterministic outputs, LLMs alone can't get the job done. But now, we have the tools to observe, measure, and test how to get there. Oh, did I mention the free to use TrustGraph Playground with full demos for OCSF and Agentic Commerce coming soon?

I'm so glad in my recent episode of Knowledge Graph Insights with Larry Swanson, that I finally got to talk about something that's been overlooked - The timeline of the explosion of the term context graph. Yes, it started with the Foundation Capital article, but what happened next? I wouldn't have even known about that article if it weren't for @KirkMarple and the response his rebuttal article got. But what was really going on? Check out my conversation with Larry for my take on why the term context graph exploded in over the holidays of 2025. Audio, video, and full transcript: knowledgegraphinsights.com/d…