Javier Loria

Hi, I’m Javier Loria. I work on three fronts that I no longer think can be separated: data architecture and engineering, data governance, and AI engineering. They’re all the same bridge between business strategy and technology execution, and I’ve spent my career on both ends of it: the first half in management (finance, sales, general management), the second in technology (development, networks, architecture). So when I model your data, I’m modeling your business, not just your tables.

I care about the parts that product demos skip: the modeling decisions, the trade-offs, the governance – who owns the data and what it means – and the details that decide whether a data solution holds up in production or quietly falls apart six months later. After 25 years doing governance work without the letters, in 2026 I sat the DAMA exams and earned the CDMP Practitioner.

AI engineering is the newest front, and the one where that history pays off most. Agents and copilots now answer business questions straight from your data, with no analyst in the middle to catch a bad definition. Whatever ambiguity lives in your semantic model, the agent repeats it with confidence. So I work on both sides of that: the agents themselves, and the governed, well-modeled data they need to be trusted. I earned the GitHub Copilot certification in October 2026, and I’m now working on AI-103 (Azure AI Apps and Agents Developer) and AB-100 (Agentic AI Business Solutions Architect), on the way to Microsoft’s Frontier Transformation Engineer badge. Yes, 35 years in, I’m studying for exams again. I’m learning parts of this as I go, like everyone. The difference is remembering which mistakes we already made with the last technology wave.

For 25 years I’ve designed and delivered data architectures on the Microsoft data platform for organizations across Latin America. Earlier in my career I spoke internationally on the Microsoft data stack with the support of SolidQ. I later founded Primus Data, a data consultancy in Costa Rica, where I work as manager and mentor, and where I’ve spent years in the trenches of real-world data projects across the region.

In the Spanish-speaking data community, I’m a known voice. This site is where I bring that work to a wider, English-speaking audience.

I’m currently writing a series of books on semantic models: the modeling foundations behind analytics that scale, aimed at practitioners who build these systems for a living.

What you’ll find here: in-depth, practical writing on semantic models, data modeling, data governance, and what changes when AI agents start reading your data; the topics I speak on and my speaking history; and, soon, the books.

Find me on LinkedIn and GitHub, or reach me at Jloria@primusdata.net.