Microsoft updates Fabric for agentic transformation
Microsoft Fabric now lets you build entire AI agent applications — data, logic, web UI, governance — without containers or external deployment. Rayfin SDK, Fabric Apps, and Osmos-powered agentic data engineering ship together.
Why it matters
Microsoft is positioning Fabric as the unified platform where enterprise teams prepare data estates, build agent-native applications, and operate them at scale. This consolidates what previously required external app platforms, data engineering tools, and observability stacks — and grounds agents in governed, contextualized data.
The key facts
8 to knowFabric Apps now available — build and deploy full applications (web pages, business logic, functions, storage) directly in Fabric without containers
Rayfin SDK (open-source) and CLI for AI-assisted development now integrated into Fabric Apps
Osmos acquisition (January 2026) brings agentic data engineering to Fabric — agents autonomously execute data transformation, validate results, refine approach with boundary-setting
Netz claims Osmos accuracy on complex data engineering significantly exceeds frontier-lab agentic systems
Monitor Hub (Observability in Fabric) provides unified monitoring across jobs, capacities, workspaces, workloads; operations agents and Activator automate root-cause analysis and response
Database Hub (public preview) — central control plane for SQL Server, Azure SQL, PostgreSQL, Cosmos DB, and Fabric SQL databases; on-premises and cloud
IQ Share — secure cross-organizational data sharing with business context, maintaining governance; agents work on same copy across orgs
Announced at European Microsoft Fabric + SQL Community Conference, Barcelona, September 29, 2026
Go to the source
CIOcio.com
Publisher excerpt: Microsoft believes the era of enterprise agentic transformation is here, underscored by a slew of Microsoft Fabric announcements at the European Microsoft Fabric + SQL Community Conference in Barcelona today. “We’re going to talk a lot about the notion of how you prepare your data estate to really…