Samuel Holcman: Thank you for tuning in to Real Talk. Be sure to join your host Sam Holcman again for another edition of our program. We'll have more Real Topics of discussion then. How do you sort out the so-called jargon from real-world practices that work? Do the members of your organization find some business or technology advice utterly confusing? Welcome to Real Talk with Sam Holcman. In this program, we set the record straight and in terms that business people and technology people can understand. Now, here is your host, Sam Holcman. From semantic hubs enterprise augmented intelligence, the missing step in agentic AI. This is Real Talk with Sam Holcman. I'm Sam Holcman. In a recent CIO article by Martin DeSolis, excuse me if I pronounced his name incorrectly, my apologies, titled How Effective Are Semantic Hubs in Moving Agentic AI Forward? The author argues that semantics are now the backbone of enterprise AI, especially as organizations rush to deploy agentic AI systems at scale. They highlight a critical shift. The challenge is no longer moving and storing data, but ensuring that the data means the same thing everywhere and however it is used. This is precisely the problem space or opportunity space that enterprise augmented information, enterprise augmented intelligence, however you'd like to pronounce that or see those initials, was designed to address. And it's a space that EACoE and BACoE have been working on for more than, yes, 50 years. Now, semantic hubs, a necessary but partial answer. The CIA article uses a powerful example from Walmart's early experimentation with a generic large language model scraping product data to generate descriptions. Because the model was not grounded in Walmart's own definition, business logic, and constraints, it produced hallucinations, fabrications, and misled content, and misleading content that hurt customer trust and conversations. The lesson is clear. Without business semantics, even the most impressive models will fail in the enterprise. To solve this, The article introduces semantic hubs as a centralized architecture that translates raw data into consistent business concepts. Instead of every application inventing its own definition of customer order or on-time delivery, a semantic hub enforces a uniform vocabulary and meaning. This becomes the backbone for more reliable, agentic AI deployments inside the walls of a single organization. Now, you know the words that's But, what's another word? However, the article also notes the inherent limitation. Enterprises do not live in isolation. They operate within sprawling ecosystems of suppliers, partners, customers, and regulators, each with their own definition and semantics. A single internal hub cannot magically resolve semantic mismatches across organizational boundaries, nor does it on its own. manage the broader issues of context and provenance that determine whether AI can be trusted. Semantic hubs are necessary. They're not sufficient. The article's quiet conclusion. You need a universal mental model. The CIA article goes further by pointing on emerging efforts like the Open Semantic Interchange Standard. OSI is described as a way to give AI systems a universal mental model for data so that agents can interpret business definition with the same precision and intent as a human expert, even across different tools and domains. In parallel, technologies like the Model Context Protocol and Agent to Agent are evolving to let AI agents talk to each other. and pull data from multiple systems safely and consistently. Put simply, the enterprise AI stack is being rebuilt around semantics, common meetings instead of isolated data, shared definitions instead of siloed reports and dashboards. Standards for all agents communicate and reason not just how APIs, application programming interfaces connect. The article closes with a clear call to action. Organizations must start planning now to ensure their data and semantics are ready for the agentic AI revolution. It frames OSI and semantic layers as the critical ingredients in that planning. Now. Let me be blunt. This is exactly where EAI enters the picture. Enterprise augmented intelligence or ⁓ enterprise augmented information as we call it, from semantics to intelligence, is not just about adding AI to existing processes. It's about designing the enterprise so that humans and AI agents can share the same governed knowledge base and the same conceptual view of the business. That requires four integrated layers. Semantics, syntax, number one. Syntax, the structured representation of the enterprise, capabilities, processes, data entities, rules, events, and relationships. Without common syntax, you cannot even talk about coherently about the business, which is the next one, semantics. The agreed meaning of those structures, what exactly is a customer, how do we define churn, what does ordership mean in each channel? This is what semantic hubs and OSI are trying to standardize. Number three, context. The conditions and situations in which meanings apply. Which regions, which product lines, which regulatory regimes, which scenarios. Agentic AI that ignores context who will be confidently wrong. And finally, provenance. The lineage of data decisions. Who defined this term? When was this rule changed? What source has fed this insight? How was this model and how has it been updated? Without provenance, there is no trust and no governance. EAI is the operating model for all four of these dimensions. They are explicit, governed, and architected. It turned semantic hubs from promising infrastructure into a foundation for enterprise-wide intelligence. And ladies and gentlemen, this is what EACoE and BACoE have been doing for 50 plus years. If you read the CIA article through an enterprise architecture and business architecture lens, you notice something that is interesting. Most of the new requirements it describes are things rigorous EACoE Enterprise Architecture and BACoE Business Architecture have been dealing with for decades. Yes, decades. Making sure that data means the same thing everywhere it's used is classic EACoE Enterprise Architecture territory. Translating between business views and systems is exactly what BACoE Business Architecture frameworks were built to do, providing a stable governed model that can outlast any specific technology wave in the core promise of serious architecture practice. EACoE for Enterprise Architecture and BACoE for Business Architecture have developed and refined methods for more than 50 years to precisely do this in real organizations long before we called it agentic AI or semantic hubs. This work has always centered on building explicit formal models of the enterprise that unifies syntax and semantics, capturing context, who does what, when, where, and under which constraints, managing provenance, how definitions, rules, and structure change over time and why. In other words, EACoE and BACoE have been constructing open semantic interchanges like Universes Inside Enterprises long before the AI community gave it that name. We're thrilled that it's happening. The current conversation about semantic layers is, in many respects, the AI world catching up with the discipline that EACAOE Enterprise Architecture has been recognized for and has been doing for decades. Now, what's the next logical step from semantic hubs to EAI? enterprise augmented intelligence is the next logical step. The CIO article correctly asserts that data and semantic readiness is now the gating factor for a successful agentic AI. The question is not whether enterprises will build semantic hubs. Many already have. The question is whether the hubs will live as isolated technical projects, which most of them are right or be integrated into a true enterprise-augmented intelligence architecture. This is where EACoE and BACoE offer a unique, immediate advantage. We provide a comprehensive blueprint for modeling syntax, semantics, contents, and governance at the enterprise level. Aligning those models with strategies, capabilities, and measurable outcomes, rather than treating the semantics as purely technical concern. This gives organizations a repeatable method to govern changes as regulation shifts, market evolves, and AI capabilities expand. So when the CIOs or CDOs or CAOs read, now is the time to make sure your data is ready for this revolution, EACoE and BACoE can credibly say we have been doing exactly this work for more than a half a century. and we have a method to help you operationalize it today. Symantec hubs and OSI are important tools. EAI is the destination. EACoE and BAC are the roadmaps to take you there. Please reach out to us at www.eacoe.org and www.bacoa.org. We can assist you in this journey. Please schedule a no-fee consultation. We are ready and able to talk to you. Thank you.