{
  "id": 10968769,
  "title": "Microsoft Fabric is where AI agents learn how the business works",
  "url": "https://urgent.news/2026/09/30/microsoft-fabric-is-where-ai-agents-learn-how-the-business-works",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-30T14:46:32.000Z",
  "source": {
    "name": "The New Stack",
    "slug": "the-new-stack",
    "url": "https://thenewstack.io/microsoft-fabric-agent-context/"
  },
  "original_language": "en",
  "account": "Microsoft Fabric, the integrated data platform, is becoming the hub where enterprise agents learn about a company's operations, whether past events or current happenings. At its FabCon and SQLCon data conference in Barcelona, Spain, Microsoft unveiled plans to simplify information gathering using Fabric. One key feature, Fabric IQ, now integrates with Microsoft 365 Copilot by default, allowing customers to query it via the MCP interface. Power BI will soon transform these definitions into apps, while ontologies, which introduce business rules, are expanding into preview. Microsoft Fabric's CTO, Amir Netz, likens agents to Drew Barrymore from 50 First Dates, emphasizing that agents forget previous information each time they open their eyes, so the first step is to inform them about their current context. Customers often discover this limitation the hard way, as corporate vice president Yitzhak Kesselman explains that advanced customers manage their agents' evaluations, while others struggle to understand the difference between expected and actual outcomes. Kesselman has interacted with over 320 companies this year, with the business side urging them to demonstrate the value and difference between \"before and after\" agent usage. Unlike coding agents, which have code, repositories, change history, specs, and tests, outside of coding, enterprises typically lack such frameworks. Thus, Fabric IQ aims to address this gap by combining unified data from OneLake, trusted metrics from Power BI semantic models, and operational context from ontologies and real-time intelligence. Storage is crucial for this process, with OneLake, Fabric's storage layer, serving as the foundation. By feeding data from various first-party and third-party services into OneLake, enterprises can provide context to their agents. Microsoft announced integrations with Salesforce Data Cloud 360, general availability of mirroring from Google BigQuery, a ClickHouse workload running directly against OneLake, and dbt's Fusion engine, soon to be introduced in Fabric. While mirroring security roles and enforcing them are not direct data copies, Microsoft assures that these permissions are actively maintained and preserved. Enterprises can also pull mirrored data out for use in other tools, as OneLake's open formats and APIs allow any engine to read and utilize it. Lastly, Microsoft envisions agents understanding not just the past and present but also the future, with semantic models, real-time intelligence, and streaming as the building blocks.",
  "summary": "Microsoft is turning Fabric, its integrated data platform, into the place enterprise agents go to learn how a company works, whether The post Microsoft Fabric is where AI agents learn how the business works appeared first on The New Stack .",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}