The Microsoft Fabric Blog introduces a preview of anomaly detection in Real-Time Intelligence (RTI). Readers learn from Microsoft Fabric Blog how this AI-powered feature helps organizations uncover data irregularities, enabling proactive monitoring and alerting in real-time business scenarios.

AI–Powered Real-Time Intelligence with Anomaly Detection in Microsoft Fabric (Preview)

A new AI capability is now available in preview for Microsoft Fabric’s Real-Time Intelligence (RTI): anomaly detection. This feature aims to empower users with scalable, proactive, AI-driven real-time data experiences, expanding what’s possible with streaming data on Microsoft Fabric.

Unlocking Value with AI in Real-Time Intelligence

Modern systems demand real-time awareness. Real-Time Intelligence (RTI) in Fabric acts as the infrastructure that connects and interprets streaming data from diverse sources, both physical and digital. The platform’s new anomaly detection feature is designed to surface ‘unknown unknowns’—subtle or unexpected data irregularities—using proven, industry-standard models.

The service brings sophisticated data science tools into reach for less technical users, enabling organizations to automate the detection of operational bottlenecks or abnormal signals and respond instantly, improving operational efficiency and reducing risk.

How Anomaly Detection Works in RTI

  1. Selecting Your Data: Start by choosing an Eventhouse in Real-Time Hub.
  2. Configuration: Select ID fields and values where you want to monitor for anomalies.
  3. Model Selection: The detector automatically tests industry-standard models to find which best fits your data’s patterns.
  4. Preview and Testing: Preview detected anomalies and try different models to see which produce the most accurate results for your use case.
  5. Alerting: Publish detected anomalies as ‘Anomaly’ events in Real-Time Hub. Alerts can be set up so users are notified via Teams messages or email when anomalies occur.

RTI Anomaly Detection Screenshot

Getting Started

To enable anomaly detection, Fabric admins need to activate the feature in the Fabric admin portal. From there, users can configure detectors directly in the Real-Time Hub and integrate automated alerts into their workflows.

For additional details, see the Real-Time Intelligence documentation.

Feedback and Support

Microsoft welcomes feedback on the preview feature. Share your experience via blog comments or the Fabric community feedback form.

This post appeared first on “Microsoft Fabric Blog”. Read the entire article here