How I created my first Azure DevOps extension for Microsoft Fabric

This post covers how I created my first Azure DevOps extension for Microsoft Fabric. Sharing my journey so that others can learn from my experience and create their own extensions.

By: Kevin Chant - MVP, Fabric Super User, Community Contributor · MVP Profile · LinkedIn · X

Native vs Custom Branching Out in Microsoft Fabric

This post covers native vs custom branching out in Microsoft Fabric, where I compare Microsoft Fabric's native branching out options with the custom Branch Out to New Workspace scripts made available by Microsoft. I decided to write this post after Microsoft introduced the new branch workspace admin configuration options in Microsoft Fabric, which are currently in preview.

By: Kevin Chant - MVP, Fabric Super User, Community Contributor · MVP Profile · LinkedIn · X

Fabric-CICD in Practice - Groundwork

First part of a series for CICD in Fabirc. Starting from an empty workspace, over three posts, we will build up to a multi-workspace solution that deploys into empty environments with nobody clicking anything. It will be opinionated, and the first opinion is that fabric-cicd is the only sensible basis for a complete CI/CD solution in Fabric. This first post is the groundwork the series will build on: the branching strategy and what each workspace is for, the sample repo, the service principals, the pipelines that promote code to higher environments, and the approval gate that protects prod, ending with a deployment into empty test and prod workspaces.

By: Jake Duddy - MVP · MVP Profile · LinkedIn

OneLake security with shortcuts and Direct Lake

This article unpacks how OneLake security, shortcuts, and Direct Lake interact to determine who can actually see data in a Fabric estate, showing that a shortcut configured as passthrough doesn't guarantee the report user's identity ever reaches the source. Using a real hub-and-spoke tenant build, it demonstrates the article's central finding: the same reader, under the same restrictive column-security role, is fully blocked under Direct Lake over OneLake but completely unrestricted under Direct Lake over a SQL analytics endpoint. It also covers two related traps, OneLake roles being Grant-only (so a narrow role can't override a broader one like DefaultReader) and workspace roles bypassing OneLake security entirely, closing with a practical checklist for testing every identity path before trusting the result.

By: Leon Gordon - MVP · MVP Profile · LinkedIn

Workspace Identity en Fabric: por qué tu pipeline no puede ejecutar el notebook aunque «todo esté bien»

A hands-on guide to Workspace Identity in Microsoft Fabric, based on a real-world case where a pipeline couldn't run a notebook even though everything looked right. It walks through the five pieces you need to configure: enabling the identity, assigning its workspace role, creating the Notebook connection, sharing connection permissions and granting access in Azure Key Vault. It also compares Workspace Identity with Service Principal, including the Semantic Link limitation. Written in Spanish.

By: Mónica Mesa - MVP · MVP Profile · LinkedIn

The Hidden Cost of the Data Swamp: Fabric Capacity Guardrails

Microsoft Fabric has fundamentally revolutionized how enterprise data teams operate. By unifying data engineering, data science, data warehousing and Power BI into a single Software-as-a-Service (SaaS) platform, it breaks down silos and lets business users move at lightning speed. However, this democratization of data comes with a massive hidden risk. When workspace creation is unchecked and pipelines run wild, speed quickly transforms into a costly “data swamp.” Unlike traditional infrastructure where compute and storage are rigidly partitioned, Fabric pools your processing power into Capacity Units (CUs). If you don’t govern those capacities, runaway queries, poorly optimized Spark jobs and unmanaged workspace sprawl will quietly drain your cloud budget before finance even sees the bill.

By: Pragati Jain - MVP, Fabric Super User, Community Contributor · MVP Profile · LinkedIn

Microsoft Fabric Migration: More Than Just Moving Data

This article provides a practical and simplified overview of Microsoft Fabric Migration, highlighting why migration should be more than simply moving data from one system to another. It covers key considerations such as Current State Assessment, Target Architecture, Medallion Architecture, Metadata-driven Pipelines, Governance, and the importance of taking a gradual approach to migration. The content is specifically aimed at the Arabic-speaking data and analytics community, helping Arabic-speaking professionals understand and apply Microsoft Fabric migration concepts in a clear and accessible way.

By: Rufyda Rahma - Fabric Super User, Community Contributor · Fabric Community Profile · LinkedIn

Liquid Clustering en Fabric, qué es, cómo configurarlo y cuándo usarlo

A practical guide (in Spanish) to Liquid Clustering in Microsoft Fabric, explaining how it compares to Hive-style partitioning and Z-Order and why its declarative approach removes much of the manual layout maintenance. It covers the differences between Runtime 1.3 and 2.0, how to declare and change clustering keys with CLUSTER BY and OPTIMIZE, and how to measure layout quality with clusteringQuality(). It closes with practical caveats before adding it to production pipelines, such as Python notebook support, V-Order compatibility and deletion vectors.

By: Mar Lizana - MVP · MVP Profile · LinkedIn

From Chaos to Clarity: A Supply Chain Response App with Fabric's SQL Database

Fabric collapses a workflow that traditionally spans several tools — a database server, an ETL tool, a BI platform, a monitoring dashboard — into one connected environment. For a supply chain disruption app specifically, that matters because disruptions don't wait for data to be reconciled across systems.

By: Rajendra Ongole - MVP, Fabric Super User · MVP Profile · LinkedIn · Social profile

Unlocking Azure Databricks Tables in Microsoft Fabric with OneLake

This article explains how to store Azure Databricks tables directly in Microsoft Fabric OneLake and make them available in Fabric through a Lakehouse. It walks through the required configuration steps in both Microsoft Fabric and Azure Databricks, including Unity Catalog and OneLake external locations, and demonstrates how to expose and reuse Databricks tables for analytics and reporting. The article also shows how data can be written from Fabric to Databricks and from Databricks back to Fabric, enabling seamless bi-directional data integration.

By: Rene Fuerstenberg - Fabric Super User · Fabric Community Profile · LinkedIn

Fabric en VSC: Desarrolla y optimiza con IA contextual — Cap. 3: Extensión SQL Server (MSSQL) para VS Code: Ahora sí podemos crear, editar y eliminar en el SQL Endpoint

EN: In Chapter 3 I fix the gap from Chapter 2: the Fabric MCP could only read the SQL Endpoint. With the MSSQL extension for VS Code and Copilot in agent mode, we can now create, edit and delete views directly. It complements the MCP rather than replacing it, and validation is still on us. ES: En este Cap. 3 resuelvo el límite del Cap. 2: el MCP de Fabric solo podía leer el SQL Endpoint. Con la extensión MSSQL para VS Code y Copilot en agent mode ya podemos crear, editar y eliminar vistas directamente. No sustituye al MCP, lo complementa, y la validación sigue siendo nuestra

By: Sara Alonso - MVP · MVP Profile · LinkedIn

Direct Lake Performance Based on Delta Layout

How you write the Delta table decides what a Direct Lake model costs to read, and the number of row groups moves it the most. Fast and cheap can be the same layout, just not one with many row groups: every group you add buys wall clock and pays for it in CU. V-Order helps on top of that, not instead of it. Measured across dozens of semantic models on one capacity, with what to write when you are optimizing for speed, for cost, or for not thinking about it, and how to control the layout without Spark.

By: Vojtěch Šíma - MVP, Fabric Super User · MVP Profile · LinkedIn

Sync Dataverse to Microsoft Fabric in minutes

Dataverse and Microsoft Fabric work brilliantly together, and setting up real-time data sync is easier than you might think. In this tutorial, I'll show you how to sync specific Dataverse tables to Microsoft Fabric in just a few clicks. This approach gives you near real-time access to your business data while avoiding the need to grant access to all Dataverse tables. Perfect for Power Platform, Power BI, and Fabric users looking for a secure and scalable way to analyse operational data.

By: Alex Ward - Fabric Super User · Fabric Community Profile · LinkedIn

Microsoft Fabric + AI: End-to-End Development Without Coding

Build a complete Microsoft Fabric data analytics solution from scratch using AI In this Microsoft Fabric End-to-End Project, I take you through the complete journey of building a modern data platform — starting with raw sales data and ending with a Direct Lake Semantic Model and Power BI report.

By: Amit Chandak - MVP, Fabric Super User · MVP Profile · LinkedIn · X

The Preview Runtime Cut My Fabric Data Agent Cost by 98%

Imagine a toggle in Microsoft Fabric that gives your data agent better answers and faster responses on hard questions. That toggle is the preview runtime. And it does one more thing: it cuts capacity utilization. Significantly, if your semantic model still carries Q&A. Learn how I saved 98% CU and in what scenarios you can gain similar savings.

By: Gil Raviv - MVP · MVP Profile · LinkedIn

Copilot Credits: Managing Consumption

The next stage of Copilot adoption, and the next step in Microsoft Investment Management is managing Copilot Credits. Based on real usage data across hundreds of clients in Prism and only requiring a simple input of the number of users in your tenant, this report will help you identify your Copilot credit requirements.

By: James Mounsey-Moran - MVP · MVP Profile · LinkedIn

Copilot y Claude en VS Code: la estructura de agentes IA para developers de Fabric y Power BI

How to structure a repository so GitHub Copilot and Claude Code work as a team of specialized AI agents for Microsoft Fabric development. It maps the .github/ and .claude/ folders side by side and explains four configuration tiers: always-on instructions, path-scoped rules, on-demand skills and deterministic scripts. It includes real SKILL.md and .agent.md examples, covers Microsoft's official skills-for-fabric repository and links to a follow-up post on day-to-day use. Written in Spanish.

By: Mónica Mesa - MVP · MVP Profile · LinkedIn

Building a Better Microsoft Fabric Data Agent: Instructions, Visual Policy, and Cost-Aware Design

How to move a Fabric Data Agent from demo to production by putting the right context at the right layer rather than writing one giant prompt. The core practice is separating Semantic Model AI Instructions (measure priority, business definitions, time rules ,what the DAX-generation tool actually uses) from Data Agent Instructions (orchestration, response and visual policy, guardrails). It treats visual output as governed adaptive responses, one visual by default, and never changing a business calculation to fit a chart and, since consumption is token-based, optimizes for cost per successful answer instead of tokens per prompt by reducing ambiguity and retries. A bonus section extends the same layered approach to SQL sources using Data Source Instructions and example queries.

By: Natarajan Manivasagan - Fabric Super User · Fabric Community Profile · LinkedIn

How does RAG actually work?

RAG is a well-defined pipeline — retrieve, format, prompt, generate, extract — built on two foundational capabilities: an external model reference and a native vector data type.

By: Parul Sagar - Fabric Super User · Fabric Community Profile

Why Your Fabric Data Agent Gives Wrong Answers (And How Data Enrichment Fixes It)

Everyone wants conversational AI on their data now — ask a question, get a number back, skip the DAX and the SQL and the three-day wait for someone to build a report. Fabric Data Agent actually pulls this off. Point it at a semantic model, a lakehouse, a warehouse, a KQL database, and it answers questions typed in plain English, and it's honestly one of the only AI features in this whole ecosystem that's lived up to its own demo. It's only as good as your metadata though, and I found that out the hard way enriching a model for a Data Agent rollout. It kept whiffing on questions that should've been trivial — same table, same data, and depending on how you phrased the question you'd get a wrong answer or nothing back at all. Took some digging to see why. The fix, once I actually saw what was going on, seemed almost stupidly simple. It's not, not really, not once you're doing it across a whole real model instead of a couple of test columns, but that's later in this post.

By: Praful Potphode - Fabric Super User · Fabric Community Profile · LinkedIn

Forecasting with Autoregression, ARIMA & AIC / BIC for Data Science in Microsoft Fabric

In this edition, we’re exploring forecasting through Autoregression, ARIMA, and the model selection tools AIC and BIC. By the time you’re done reading, you’ll understand how data can actually learn from its own patterns, how ARIMA helps bring structure to unpredictable trends, and how AIC and BIC keep your models grounded by balancing accuracy with simplicity.

By: Sahir Maharaj - MVP · MVP Profile · LinkedIn

Model Explainability That Goes Beyond the Notebook for Data Science in Microsoft Fabric

In this edition, we’re exploring model explainability in production and how SHAP, FastAPI, and Plotly can work together to make predictions easier to understand. By the time you’re done with this, you’ll have a clear view of how SHAP can explain individual model decisions, how those explanations can be served dynamically through an API, and how interactive visualizations can turn technical attribution values into something far more approachable.

By: Sahir Maharaj - MVP · MVP Profile · LinkedIn

Create a Fabric app with write-back

In this article, I will guide you through building a Fabric App from scratch, step by step. I will explain the requirements, detail how I built the app, and share my experience using an AI coding agent instead of doing everything manually. I also recorded a video about the process, which you can find in the blog post.

By: Valerie Junk - MVP · MVP Profile · LinkedIn · YouTube

Kurt Buhler - discussions on AI | Explicit Measures (563–566)

Guest Kurt Buhler joins Explicit Measures for episodes 563–566 covering how Fabric apps and AI fits into the Power BI world, self-awareness in data and AI literacy, identity and Power BI, and agentic development of Power BI reports. This public playlist collects those four full episodes plus the published YouTube Shorts clipped from the Kurt episodes.

By: Mike Carlo - MVP · MVP Profile · LinkedIn

Global Aircraft✈️ Live Tracking with Microsoft Fabric Real-Time Intelligence

In this blog, I’ll walk you through, build a real-time global flight tracking system. We will ingest live flight data from the public OpenSky Network API using a Python polling script, stream it through Microsoft Fabric Eventstream into an Eventhouse (KQL Database), transform the dense raw arrays using KQL update policies and visualize the results on a Real-Time Dashboard complete with maps, KPIs and analytical charts

By: Inturi Suparna Babu - Fabric Super User · Fabric Community Profile · LinkedIn

Inventory Is a Status, Not a Sum

Inventory is a point-in-time status, so adding up totals can hide the real picture. This article shows how to model inventory in Power BI by storing the latest stock level once as a reference and keeping the transactions with reversed signs, then writing DAX that returns stock at any date. On a sample retail dataset, calculating per store and product shows negative balances, lines at risk of running out within 90 days, and slow-moving stock, all hidden behind a healthy-looking total. It also covers forecasting stock from the sales run rate and valuing inventory likely to sit unsold for over a year.

By: Ilgar Zarbaliyev - MVP, Fabric Super User, Community Contributor · MVP Profile · LinkedIn · X

Semantic Models in Microsoft Fabric: Building a Trusted Business Layer for Power BI

Semantic models provide the business layer between data and Power BI reports in Microsoft Fabric. This article explains how semantic models define relationships, measures, business logic, security, and reusable metrics so teams can work from consistent definitions. It also covers how semantic models connect with Lakehouses, Warehouses, OneLake, Direct Lake, and Power BI to create a trusted foundation for analytics.

By: Olufemi Olamoyegun - Fabric Super User, Community Contributor · Fabric Community Profile · LinkedIn

Who Can See What? A Complete Power BI Access Audit Solution in Microsoft Fabric

Fabric notebooks harvest RLS/OLS definitions, role memberships, AD group members, and workspace roles into a lakehouse and a Power BI report on top answers every "who can see what?" question on demand.

By: Praful Potphode - Fabric Super User · Fabric Community Profile · LinkedIn

La importancia de un buen modelo de datos (más que nunca, ahora que hay IA de por medio)

English: AI doesn't let us skip good data modeling: without a solid model (clear relationships, the right granularity, a star schema), Copilot and agents produce DAX that "sort of" works until it breaks. I share what I do before writing any DAX and how to diagnose an inherited model with Performance Analyzer and DAX Studio. Español: La IA no nos libra de modelar bien: sin un modelo sólido (relaciones claras, granularidad correcta, esquema en estrella) Copilot y los agentes generan DAX que "más o menos" funciona hasta que falla. Repaso qué hago antes de escribir DAX y cómo diagnosticar un modelo heredado con Performance Analyzer y DAX Studio.

By: Sara Alonso - MVP · MVP Profile · LinkedIn

Will Fabric Apps Replace Power BI? Fabric Now- OLTP + OLAP

In this video, I explore how to create a Microsoft Fabric App using the Rayfin SDK and demonstrate how modern development tools can be used to build interactive applications directly around your Fabric data. In this video, we explore: What are Microsoft Fabric Apps? Introduction to the Rayfin SDK Exploring the Fabric App user interface Developing a Fabric App using Visual Studio Code Using GitHub Copilot for AI-assisted development Building a reporting experience inside a Fabric App Creating a Data Entry Sales Application Adding a Semantic Model experience to the same application And finally... asking the big question: Is Microsoft Fabric becoming more than a Data Analytics platform? And... Is Power BI Dead? 👀

By: Amit Chandak - MVP, Fabric Super User · MVP Profile · LinkedIn · X