From Data to Knowledge: Building AI-Ready Semantic Models with Ontologies in Microsoft Fabric

Microsoft Fabric Data Agents can query your semantic models, but do they truly understand your business?

Traditional semantic models are excellent at organizing data for reporting and analytics. AI, however, needs more than tables, measures, and relationships. It needs business meaning.

In this session, you'll discover how semantic models and ontologies work together to create a knowledge layer for Microsoft Fabric. By adding shared business definitions, context, and relationships, ontologies help Data Agents reduce ambiguity, reason more effectively, and deliver more trustworthy answers.

Through practical demonstrations, we'll show how business knowledge can enrich semantic models and enable AI experiences that understand not just your data, but the business concepts behind it.

Key takeaways

Why AI needs business context, not just data structures
How ontologies help Data Agents understand business concepts and relationships
How semantic models and ontologies complement each other
How to build more trustworthy and scalable AI solutions in Fabric

Bottom line:
Semantic models help AI find data.
Ontologies help AI understand the business. Together, they create AI-ready analytics.

Audience & prerequisites
Whether you're a Fabric architect, BI professional, data engineer, or AI enthusiast, You'll leave with practical patterns, architectural guidance, and techincal examples for using Semantic Models, Ontologies to empower data agents.

Create Once, Use Everywhere: A Standard Date Dimension in Fabric

A date dimension is one of the foundations of every data platform. In this 5-minute session, I'll show how to build, enrich, and store a reusable date dimension in your Fabric Lakehouse.

Why?
• Consistency across all reports and models.
• Standardization drives automation.
• Build once, use everywhere.

Key takeaway:
Create a single calendar that powers your entire Fabric platform.

Transforming GEO Data from Tables to Fabric Maps

Did you know you can turn location data into interactive maps in Microsoft Fabric?
In this 5-minute session, I'll show how to transform geographic data into GeoJSON and visualize it with Fabric Maps.
Why?
Location data tells a story. Maps make it visible.
Turn location data into a powerful decision-making tool.

Key takeaway:
Learn how a few simple transformations can bring your location data to life with Fabric Maps.

Create Field Parameters Automatically with TOM for Power BI

Field Parameters make Power BI reports more flexible, but creating and maintaining them manually doesn't scale. I'll show how to generate Field Parameters programmatically using the Tabular Object Model (TOM) from a Fabric notebook.
Why does this matter?

Automate repetitive configuration tasks. and reuse the same logic across multiple semantic models.

Takeaway: Treat your semantic model as code and eliminate manual Field Parameter maintenance.

Medallion Magic: Turning Data Layers into Stories That Stick

Ever wondered why every data platform follows a hidden pattern? Meet the Medallion Architecture: the secret behind all data architectures. Do you remember classic BI solutions? Or data layers like Cleansed or Curated?
In this session, discover why you’ve been using it all along, how to apply it intentionally, and turn everyday about your data platform into stories that connect and inspire.
Based on real‑world experience, this session will help you:
Recognize the Medallion pattern in both modern data platforms and classic BI architectures
Understand the business purpose of Bronze, Silver, and Gold layers
Translate technical data layers into clear business value
Use the Medallion Architecture as a storytelling framework to align business and IT stakeholders
By the end of the session, you’ll see why the Medallion Architecture feels so familiar, and find the secret behind all data architectures.
What more could you ask for? Stick to the plan, and you will get there!

This session is designed for a broad, business‑oriented audience:
Data Engineers, Business Analysts, and Managers, anyone involved in data decisions who wants conceptual clarity without deep technical detail.
Prerequisites:
No prior knowledge of Medallion Architecture is required.
Complexity: 100 (Business - Overview) / 200 (Intermediate)

Model Smarter, Not Harder: Build Semantic Models as Products

Building a semantic model should not feel like starting from scratch every time. Yet many organizations still rely on manual processes that are difficult to maintain, scale, and recover when changes are made.
In this session, you'll discover how Microsoft Fabric enables a more standardized approach to semantic models, helping teams reduce repetitive work, improve consistency, and accelerate the delivery of trusted business insights.
Rather than focusing on individual models, we'll explore how semantic models can be treated as reusable and scalable assets. You'll learn why metadata, naming conventions, standardized date tables, and automation are becoming increasingly important in a world of AI-powered analytics and Data Agents.

Semantic models should become repeatable, maintainable, and scalable assets instead of manually crafted one-off solutions.

During this session you will see the following benefits.
• Understand why standardization is essential for scalable semantic models.
• Learn how metadata improves discoverability, governance, and AI readiness.
• See how automation reduces maintenance effort and increases consistency.
• Discover how semantic models can evolve from one-off solutions into reusable business assets.

Audience: Data professionals, BI developers, analytics engineers, and Fabric practitioners interested in scaling semantic models.
Prerequisites: Basic knowledge of semantic models and Power BI/Fabric concepts.