Scenario 1: Sales Volume Alert

Publisher: Notebook | Consumer: Activator

Business context

A retail company runs daily sales processing in Microsoft Fabric. Each store is expected to process a baseline number of transactions per hour based on historical data. If a store's transaction volume drops significantly below its expected baseline, it may indicate a failure in the point-of-sale integration, not a natural slowdown.

The operations team needs to be notified automatically when this condition is detected, without relying on a human to monitor dashboards.

The problem without Business Events: The notebook that detects the condition would need to call a notification service directly, creating a hard dependency. If the notification service changes, the notebook code must change too.

The solution with Business Events: The notebook publishes a Retail.Sales.VolumeAlert event. Activator subscribes to that event and triggers the notification. The notebook does not know or care how the alert is delivered.

Architecture

flowchart LR
    subgraph Fabric Notebook
        A[Detect volume drop\nper store per hour]
    end

    subgraph Real-Time Hub
        B([Retail.Sales.VolumeAlert\nBusiness Event])
    end

    subgraph Activator
        C{Evaluate rule:\ndeviation_pct < -30}
        D[Send Teams alert\nto ops team]
    end

    A -->|Publish event| B
    B -->|Trigger| C
    C -->|severity = high| D

Step 1: Create the Business Event

Before publishing any event, you need to define it in Real-Time Hub.

Go to Real-Time Hub → Business Events → Create.
Create or select an Event Schema Set. Use RetailSales as the schema set name. Note the name and workspace. They map directly to SCHEMA_SET_NAME and WORKSPACE_NAME in the publisher code.
Name the event Retail.Sales.VolumeAlert.
In the schema editor, paste the following JSON and select Create:

{
  "type": "record",
  "name": "Retail.Sales.VolumeAlert",
  "fields": [
    {
      "name": "store_id",
      "type": "string",
      "doc": "Unique identifier of the store reporting the alert"
    },
    {
      "name": "expected_transactions",
      "type": "int",
      "doc": "Expected number of transactions based on historical average"
    },
    {
      "name": "actual_transactions",
      "type": "int",
      "doc": "Actual transactions recorded in the current monitoring window"
    },
    {
      "name": "deviation_pct",
      "type": "float",
      "doc": "Percentage deviation from expected volume. Negative means below expected."
    },
    {
      "name": "window_start",
      "type": "string",
      "doc": "Start of the monitoring window in ISO 8601 format"
    },
    {
      "name": "severity",
      "type": "string",
      "doc": "Alert severity level: low, medium, high"
    }
  ]
}

Step 2: Publisher - Notebook

The notebook queries the sales data, calculates the deviation per store, and publishes a Business Event for each store that exceeds the threshold.

Create the Notebook

In your Fabric workspace, select + New item and create a Notebook.
Attach it to a Lakehouse that contains your sales data, or use the hardcoded values provided below for testing.

Notebook code

Replace WORKSPACE_NAME with your Fabric workspace name and SCHEMA_SET_NAME with the name of the Event Schema Set associated with the Business Event you created in Real-Time Hub → Business Events.

# Publish a Retail.Sales.VolumeAlert Business Event from a Fabric Notebook
# notebookutils reference: https://learn.microsoft.com/en-us/fabric/data-engineering/microsoft-spark-utilities

WORKSPACE_NAME = "<your-workspace-name>"
SCHEMA_SET_NAME = "<your-schema-set-name>"
EVENT_NAME = "Retail.Sales.VolumeAlert"

# Calculate deviation for a store
store_id = "store-mx-042"
expected = 1250
actual = 743
deviation_pct = round(((actual - expected) / expected) * 100, 2)
severity = "high" if deviation_pct < -30 else "medium" if deviation_pct < -15 else "low"

# Build the event payload
payload = {
    "store_id": store_id,
    "expected_transactions": expected,
    "actual_transactions": actual,
    "deviation_pct": deviation_pct,
    "window_start": "2024-06-22T09:00:00Z",
    "severity": severity
}

# Publish to Business Events
notebookutils.businessEvents.publish(
    eventSchemaSetWorkspace=WORKSPACE_NAME,
    eventSchemaSet=SCHEMA_SET_NAME,
    eventTypeName=EVENT_NAME,
    eventData=payload,
    dataVersion="v1"
)

Step 3: Consumer - Activator

When the Retail.Sales.VolumeAlert Business Event is published to Real-Time Hub, Activator can subscribe to it and trigger an action when the rule condition is met.

Set up the alert rule

In Microsoft Fabric, navigate to Real-Time on the left navigation bar.
Select Business events in Real-Time Hub.
Locate the Retail.Sales.VolumeAlert event under your schema set.
Select Set alert from the event options.

Configure the rule

Details section

Enter a name for the rule, for example: Sales Volume Drop Alert.

Monitor section

In the Source field, select Business events.
In the Connect data source wizard, select the Retail.Sales.VolumeAlert event type.
Select + Filter and configure the following filter:
- Field: deviation_pct
- Operator: Less than
- Value: -30
Select Next, review the settings, and select Save.

Condition section

Set Check to On each event.

Action section

For Select action, choose Teams then Channel post.
Configure the Teams channel to notify the operations team.
In Context, add store_id, actual_transactions, and deviation_pct so the alert includes the relevant details.

Step 4: End-to-end test

Once you have the Business Event defined, the notebook configured, and the Activator rule in place, run the notebook as-is. The hardcoded values for store-mx-042 produce a deviation_pct of -40.56, which exceeds the -30 threshold and should trigger the Teams alert.

If the alert fires, your end-to-end setup is working. You can then replace the hardcoded values with real sales data from your Fabric lakehouse or warehouse.

What happens next

When Activator receives the Retail.Sales.VolumeAlert event and the rule condition is met, it can trigger one or more actions. This scenario uses a Teams alert, but Activator supports additional action types that you can combine or substitute.

flowchart LR
    BE([Retail.Sales.VolumeAlert]) -->|rule match| ACT[Activator]
    ACT --> TEAMS[Teams / Email alert]
    ACT --> PA[Power Automate flow]
    ACT --> NB[Notebook]
    ACT --> DP[Data Pipeline]
    ACT --> UDF[User Data Function]

Action	When to use
Teams / Email alert	Notify a team or individual immediately when the condition is met
Power Automate flow	Trigger a business process or approval workflow outside Fabric
Notebook	Run a remediation or deeper analysis in response to the event
Data Pipeline	Kick off a downstream data processing job
User Data Function	Call custom logic or an external API

For full details on configuring each action type, see the Activator documentation.