Skip to content

Scenarios

Each scenario is a complete, end-to-end walkthrough of a real business problem solved with Business Events. Every scenario includes a business context, an architecture diagram, an Avro event schema, publisher code, and consumer configuration.

Detect & Alert

Use this pattern when you want to notify a person or system the moment a condition is met. The publisher detects a condition and emits an event. Activator evaluates a rule and immediately notifies a team or triggers a response. These scenarios are the fastest path from a condition to an action.

Scenario Industry Publisher Consumer Level
Sales Volume Alert Retail Notebook Activator Beginner
Real-Time Stream Alert Finance Eventstream Activator Intermediate

Good first scenario

Start with Sales Volume Alert if you are new to Business Events. It covers the full publish-subscribe cycle with the minimum number of moving parts.

Decide & Act

Use this pattern when you want one event to trigger multiple automated actions across independent systems. The publisher evaluates context — customer history, fraud rules, inventory levels, SLA data — makes a decision, and publishes an event that carries the outcome. Multiple independent consumers act on it simultaneously without knowing about each other. Adding a new action means adding a new subscription, not changing the publisher.

Scenario Industry Publisher Consumer Level
Customer Loyalty Milestone Retail Activator Activator Intermediate
Abandoned Cart Recovery Ecommerce User Data Function Activator, Eventhouse Intermediate
Fraud Response Before Fulfillment Finance User Data Function Activator, Eventhouse Advanced
Real-Time Viewer Recommendations Media User Data Function Activator, Eventhouse Advanced
Shipment Delay Auto-Mitigation Logistics User Data Function Activator, Eventhouse Advanced

The fan-out advantage

Fraud Response Before Fulfillment and Shipment Delay Auto-Mitigation each trigger three independent Activator rules from a single event. This is the fan-out pattern: one publisher, many independent consumers, zero coordination overhead.

Track & Analyze

Use this pattern when you want every event stored as a queryable, auditable record. Every event is automatically persisted in Eventhouse as a structured, queryable record. These scenarios show how to build operational dashboards, audit trails, and reorder pipelines using KQL — without writing a single line of ingestion code.

Scenario Industry Publisher Consumer Level
Low Stock Threshold Supply Chain User Data Function Eventhouse Beginner
Data Lineage Audit DataOps Notebook Eventhouse Intermediate

Combine patterns

Most production solutions combine patterns. Abandoned Cart Recovery, Fraud Response Before Fulfillment, Real-Time Viewer Recommendations, and Shipment Delay Auto-Mitigation pair Decide & Act with Track & Analyze: Activator takes the immediate action while Eventhouse stores the event for reporting, model retraining, or compliance.

How to read a scenario

Each scenario follows the same structure:

  1. Business context — the real-world problem and why it matters
  2. Architecture — Mermaid flow diagram of the full solution
  3. Step 1: Create the Business Event — Avro schema definition
  4. Step 2: Publish from the source workload — publisher code (Notebook, User Data Function, Eventstream, or Activator)
  5. Step 3: Configure the consumer — Activator rule setup and/or KQL queries for Eventhouse
  6. Step 4: Test end to end — how to verify the full flow
  7. What happens next — extension ideas and related scenarios

Ask in our community

Join developers discussing Microsoft Fabric on Reddit.
Not a member yet? Join r/MicrosoftFabric.