Fabric Fraud Intelligence - Agentic fraud investigation on Microsoft Fabric
Fabric Fraud Intelligence is an end-to-end fraud detection and investigation demo built on Microsoft Fabric. It combines a Rayfin, Fabric App, Fabric SQL, Lakehouse, Fabric IQ Ontology, Data Agent, Eventhouse/KQL, Power BI, and an agentic investigation workflow for card fraud, AML, claims fraud, entity graphs, and human-approved decisions. Key highlights: Fraud command center with alerts, risk scoring, and exposure KPIs Case detail with Customer 360, evidence, timeline, and grounded agent chat Entity graph to explore fraud typologies and connected customers Fraud IQ scenario showing how investigation can move from manual multi-step work to an explainable agentic workflow Human-in-the-loop decisions with auditability and OneLake writeback patterns The public deployment uses synthetic seed data and deterministic agent responses. It does not expose anonymous access to the underlying Fabric SQL database, Lakehouse, Ontology, Data Agent, Eventhouse, semantic model, or report. VibeHub page: https://vibehub.microsoft.com/app/etienne-sigwald-fabric-fraud-intelligence GitHub repository: https://github.com/EtienneSIG/Fabric_Fraud_analysis Demo video: https://raw.githubusercontent.com/EtienneSIG/Fabric_Fraud_analysis/main/video/Fabric%20Fraud%20Intelligence%20%28live%29.mp4Fabric Fraud Intelligence — Agentic fraud investigation on Microsoft Fabric
Fabric Fraud Intelligence is an end-to-end fraud detection and investigation demo built on Microsoft Fabric. It combines a Rayfin Fabric App, Fabric SQL, Lakehouse, Fabric IQ Ontology, Data Agent, Eventhouse/KQL, Power BI, and an agentic investigation workflow for card fraud, AML, claims fraud, entity graphs, and human-approved decisions. Key highlights: • Fraud command center with alerts, risk scoring, and exposure KPIs • Case detail with Customer 360, evidence, timeline, and grounded agent chat • Entity graph to explore fraud typologies and connected customers • Fraud IQ scenario showing how investigation can move from manual multi-step work to an explainable agentic workflow • Human-in-the-loop decisions with auditability and OneLake writeback patterns The public deployment uses synthetic seed data and deterministic agent responses. It does not expose anonymous access to the underlying Fabric SQL database, Lakehouse, Ontology, Data Agent, Eventhouse, semantic model, or report. VibeHub page: https://vibehub.microsoft.com/app/etienne-sigwald-fabric-fraud-intelligence GitHub repository: https://github.com/EtienneSIG/Fabric_Fraud_analysis Demo video: https://raw.githubusercontent.com/EtienneSIG/Fabric_Fraud_analysis/main/video/Fabric%20Fraud%20Intelligence%20%28live%29.mp4
Building a Modern Marketing Campaign Operations App with Microsoft Fabric & Rayfin
Excited to share a new project: Fabric App – Rayfin Marketing Campaign Operations! Built using Microsoft Fabric, Rayfin, React, interactive maps, persistent workflows, and performance analytics, this application demonstrates how organizations can move beyond dashboards and create operational apps directly on top of Fabric data. ✅ Campaign workflow management ✅ Interactive geospatial visualisation ✅ Embedded analytics & KPIs ✅ Modern React-based experience A great example of how Microsoft Fabric can power both analytics and business applications on a unified platform. Marketing teams often manage campaigns across multiple regions, channels, and vendors. While analytics platforms provide visibility into campaign performance, operational actions still happen outside the analytics environment through emails, spreadsheets, and disconnected applications. This creates challenges such as: Limited operational visibility Manual campaign tracking Delayed decision-making Fragmented workflows Lack of location-based campaign insights The application combines: Microsoft Fabric as the unified data platform Rayfin for application development React for modern user interaction Geospatial Maps for regional campaign visibility Persistent Workflows for campaign management Analytics Dashboards for performance monitoring The solution demonstrates how organizations can: Business Need Solution Campaign Tracking Persistent workflow management Regional Insights Interactive maps Performance Monitoring Embedded analytics Faster Decisions Real-time data access Operational Efficiency Unified application experience Conclusion: This project demonstrates how Microsoft Fabric and Rayfin can be combined to create modern, data-driven operational applications that move beyond traditional dashboards. By integrating interactive maps, persistent workflows, React-based experiences, and analytics, organizations can empower business users to both understand and act on their data from a single application. If you are exploring how to operationalize your Fabric investments beyond reporting, this project provides a practical example of what's possible.Building a Modern Marketing Campaign Operations App with Microsoft Fabric & Rayfin
🚀 Excited to share my latest Microsoft Fabric application built using Rayfin, React, interactive maps, persistent workflows, and performance analytics. This solution demonstrates how Microsoft Fabric can be used for more than analytics by enabling operational business applications directly on top of Fabric data. I recently built an interactive marketing campaign operations application using Microsoft Fabric, Rayfin, React, interactive maps, persistent workflows, and real-time analytics. The goal was simple: demonstrate how business teams can move beyond static reports and spreadsheets to a fully interactive operational application running directly on top of Fabric data. The Challenge Marketing teams often manage campaigns across multiple regions, channels, and vendors. While analytics platforms provide visibility into campaign performance, operational actions still happen outside the analytics environment through emails, spreadsheets, and disconnected applications. This creates challenges such as: Limited operational visibility Manual campaign tracking Delayed decision-making Fragmented workflows Lack of location-based campaign insights Solution Architecture The application combines: Microsoft Fabric as the unified data platform Rayfin for application development React for modern user interaction Geospatial Maps for regional campaign visibility Persistent Workflows for campaign management Analytics Dashboards for performance monitoring Key Features 🗺️ Interactive Campaign Maps One of the most exciting capabilities is geospatial visualization. Users can: View active campaigns by location Identify regional performance differences Drill into campaign details Monitor coverage across territories Instead of looking at rows in a table, stakeholders can instantly understand campaign reach on a map. 🔄 Persistent Workflow Management Campaign execution requires multiple steps and stakeholders. The application enables: Campaign creation Approval workflows Status tracking Ownership management Execution monitoring Workflow state persists even after users leave and return to the application, creating a true operational experience rather than a simple dashboard. 📊 Campaign Performance Analytics Analytics are embedded directly into operational processes. Business users can monitor: Campaign ROI Regional performance Campaign effectiveness Operational KPIs Trend analysis This removes the gap between insights and actions. ⚛️ Modern React-Based Experience The application delivers a responsive and intuitive experience using React. Benefits include: Fast performance Interactive user experience Component-driven architecture Easy extensibility Mobile-friendly design Why Rayfin + Microsoft Fabric? Many organisations already use Fabric for analytics but struggle when they need operational applications. Rayfin enables teams to: ✅ Build applications directly on Fabric data ✅ Reduce application development complexity ✅ Accelerate business solution delivery ✅ Combine analytics and operational workflows ✅ Create engaging user experiences without extensive custom development This dramatically shortens the time between identifying a business problem and delivering a working solution. Repository 🔗 GitHub: https://github.com/nikunj11itdhm/Fabric-App-Rayfin-MarketingCampaign I'd love to hear feedback from the community and discuss ideas for future enhancements such as AI-powered recommendations, Copilot integration, and predictive analytics. #MicrosoftFabric #FabricApps #Rayfin #React #PowerBI #DataEngineering #Analytics #DataApps #MicrosoftDataPlatform #CommunityShowcaseFactory control center: image triage to annotate near incidents
This demonstration application shows how Fabric apps can be used in conjunction with Azure IoT Operations. On the edge of a local network, of a factory, building, or vessel, Azure IoT Operations can work with local devices. It can send data to Microsoft Fabric Real-Time Intelligence conditionally, using a secure connection via Azure Arc. What if we use a vibe-coded Fabric App for triage on images, taken at the factory floor while a vibration sensor detects abnormal behavior? On the Edge, an IP camera video stream and a LoRaWAN vibration sensor telemetry stream are combined using custom logic running in a Kubernetes pod. That data is then sent to an Eventstream custom endpoint and forwarded to an Eventhouse KQL Database table. Using table update policies, the table rows are cleaned and exposed in a Fabric Semantic model. Using vibe-coding in Visual Studio Code together with GitHub Copilot, the Fabric app is created. It shows how the latest images with the abnormal vibration deviation are presented. The user can then do triage and annotate the reason. This can then be shared with colleagues, customer or used for training agents. The application is built in ~1 day and demonstrates various features. Check this blog post. You can read about the full story on how the demonstration solution is built. Source of the custom code is available on GitHub.Fabric Discord Contest
🎯 New in the Microsoft Fabric Discord community: Certification Prep Challenge The idea: build something that helps you (or someone else) study for a Microsoft Fabric certification. Doesn't matter which one - DP-600, DP-700, whatever you're working toward or have your eye on. Designed to take 1-2 hours. If you get into it and want to spend more time, go for it, but you don't have to. Participants have the opportunity to get vouchers for DP-700 or DP-600 exams! Some ideas to get you started: • A flashcard site for key terms and concepts (SKUs, CU math, roles, whatever trips you up) • A quiz app pulling from exam objective areas • A "spot the error" tool: show a broken config or a wrong answer and have people pick what's off • A progress tracker across exam domains • Anything interactive that helps the content stick Two ways to enter: 📸 Concept track: no Fabric environment? No problem. Mock up your idea in Canva, Figma, PowerPoint, whatever you've got. 🛠️ Builder track: actually build it. A great excuse to try Fabric Apps and Rayfin, both in public preview right now. Submit a live URL plus a GitHub repo. To submit: Join the Discord Server and share your app or mockup in our contest entries channel. 📅 Opens: August 18 📅 Closes: September 1 🏆 Prizes: winners in each track take home some Microsoft swag Join the Discord server to enter today! https://discord.gg/EGD4Yv8DEZMetadata Driven Framework
Git Hub Code Repo : https://github.com/NandanHegde15/MSFTFabric-Projects/blob/main/Metadata%20Driven%20Framework/README.md Demo : https://www.youtube.com/watch?v=Z7mjHmX8-40 Metadata Driven Framework is a Microsoft Fabric-based end-to-end healthcare analytics platform . This project demonstrates modern data engineering practices including: Multi-layer data architecture (Raw → Silver → Gold layers) FHIR healthcare data modeling with support for patients, claims, coverage, and organizations Automated data pipelines for ingestion, transformation, and aggregation Power BI semantic models for enterprise analytics and reporting Row-level security for data governance and privacy AI-powered insights using Copilot & Data Agent capabilities Authors : Hemanth Kotha, Nandan Hegde https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DZ7mjHmX8-40Medallion Architecture in Microsoft Fabric: Building an AI-Ready Data Pipeline
Video Walkthrough Link------------> (https://raw.githubusercontent.com/trivedisunita/Fabric-Data-Factory-Contest/main/Fabric%20Short%20Video2.mp4) This project presents an end-to-end data pipeline built for the Data Factory Challenge using Microsoft Fabric. The objective was to transform raw job salary data into a structured, analytics-ready format for insights. ------------- Architecture Overview-------- Step1- ##Bronze Layer:-- Raw data is ingested into a Lakehouse from a SharePoint source without transformations, ensuring data is preserved in its original format. Step 2-- ##Silver Layer:-- Data is processed using Data Factory pipelines and stored in a Data Warehouse schema. Transformations include data cleaning, handling null values, and standardizing fields. SQL scripts and Copilot were used to assist in transformations. Step--3 ##Metadata Pipeline:--- Metadata-driven pipeline is implemented to manage and monitor data flow across layers. It enables dynamic execution, parameter handling, and improves pipeline scalability and maintainability within Microsoft Fabric. Step--4 ##Gold Layer:-- Data is modeled using dbt to create fact and dimension tables. A star schema is built for efficient querying, with reusable transformations such as converting coded fields into readable formats. Step--5 ##Semantic Model & Reporting:-- An AI-ready semantic model is created and used in Power BI. The model is optimized by selecting relevant columns and defining structure for better performance and accurate insights. Interactive reports are built on top of this model, enabling natural language queries using Copilot and delivering reliable business insights. https%3A%2F%2Fyoutu.be%2FskgWJyBPMcsDiscord Ontology
This ontology models the Fabric Community Discord Server and how it operates. It captures the key concepts of members, access control, communication, and moderation that govern the day to day server life. Users join and become members, earn roles, access channels, post messages, and if they're naughty, get dealt with through moderation actions. Rather than modelling Discord's underlying infrastructure, the focus is on the operational system running on top of it: who belongs, what they're allowed to do, how they communicate, and how the community keeps itself in check. The relationships tell a story about how access, identity, and behaviour interlock. A member earns a role, typically through a bot, and that role grants access to specific channels. Inside those channels, members post messages. Messages can then be turned into threads for deeper conversation, or they can collect reactions. Bots monitor messages and can respond if required. When someone steps out of line, a moderation action gets issued, typically automatically by a bot. Roles can get revoked, and access gets restricted, or members can be removed from the community entirely. It's a feedback loop between participation and permission, and it's the kind of structured reasoning an AI agent could actually use to answer questions like "What can this member see?", "Who moderated this user and why?", or "Which channels are open without a verified role?" The Ontology can be found here: https://github.com/taylorsamy/Ontology-Playground/blob/main/catalogue/community/tayloramy/discord_ontology/discord-server-ontology.rdf https://microsoft.github.io/Ontology-Playground/#/share/
