# Create alerts on real-time metrics in Amazon Connect
<a name="rule-real-time-metrics"></a>

You can create rules that automatically send emails or tasks to managers based on the values of real-time metrics. This enables you to alert managers on contact center operations that could potentially impact the end-customer experience. For example, you can set up an alert that sends an email to a manager when one or more agents on their team have been on break for longer than 30 minutes.

**Topics**
+ [Step 1: Define rule conditions](#conditions-rtm)
+ [Step 2: Define rule actions](#rule-actions-rtm2)

## Step 1: Define rule conditions
<a name="conditions-rtm"></a>

1. On the navigation menu, choose **Analytics and optimization**, **Rules**.

1. Select **Create a rule**, **Real-time metrics**.

1. Under **When**, use the dropdown list to choose from the following event sources: **There is an update in queue metrics**, **There is in update in routing profile metrics**, ** There is an update in agent metrics**, and **There is an update in flow metrics**. These options are shown in the following image.   
![\[The option When a real time metric is available.\]](http://docs.aws.amazon.com/connect/latest/adminguide/images/rtm-rule-condition.png)

1. Choose **Add condition**. The **Metrics** card is added automatically, as shown in the following image.   
![\[The condition for when a real-time metric is updated.\]](http://docs.aws.amazon.com/connect/latest/adminguide/images/rtm-rule-condition-all.png)
**Note**  
You can add up to 2 Metrics cards. This enables you to create a condition where one card evaluates real-time metrics and another evaluates trailing windows of time. For example, you may want an alert when several agents on are lunch break (Agent activity = Lunch break for 1 hour) and Average handle time is greater than 5 minutes.
You can add up to 10 metrics to each **Metrics** card.

   Following are the available real-time metrics you can add, depending on the event source. 
   + **There is an update in queue metrics - real-time**
     + [Contacts in queue](metrics-definitions.md#contacts-in-queue): Build rules that run when the number of contacts in a queue is a specified value. 
     + [Oldest contact age](metrics-definitions.md#oldest-real-time): Build rules that run when the oldest contact in queue reaches a specified age. 
     + [Agents available](metrics-definitions.md#available-real-time): Build rules that run when the number of agents available to handle contacts reaches a specified value. 

     The following image shows a condition that is met when **Contacts in queue** is greater than or equal to 400 AND **Oldest contact agent** is greater than or equal to 10 minutes AND **Agents available** is greater than or equal to 0, for the **Basic Routing Profile**.   
![\[Multiple real time metrics in a condition.\]](http://docs.aws.amazon.com/connect/latest/adminguide/images/rtm-rule-condition-example.png)

     To evaluate the condition with OR instead of AND, change the **Logic** setting to **Any**. 
   + **There is an update in queue metrics - trailing windows of time**

     Trailing windows of time is the past x minutes or hours. 
     + [Average handle time](metrics-definitions.md#average-handle-time): Build rules that run when the average handle time reaches a specified duration. 
     + [Average queue answer time](metrics-definitions.md#average-queue-answer-time): Build rules that run when the average queue answer time reaches a specified duration. 
     + [Average agent interaction time](metrics-definitions.md#average-agent-interaction-time): Build rules that run when the average interaction time reaches a specified duration. 
     + [Average customer hold time](metrics-definitions.md#average-customer-hold-time): Build rules that run when the average hold time reaches a specified duration. This metric does not apply to tasks so the value for them is always 0.
     + [Service level](metrics-definitions.md#service-level): Build rules that run when the service level reaches a specified percent. 
   + **There is in update in routing profile metrics**

     Trailing windows is not available for rules based on routing profiles.
     + [Agents available](metrics-definitions.md#available-real-time): Build rules that run when the number of agents available to take inbound contacts reaches a specified value. 
   + **There is an update in agent metrics - real-time**
     + [Agent activity](metrics-definitions.md#agent-activity-state-real-time): Build rules that run when the agent activity equals a certain value such as Available, Incoming, On contact, and more.
   + ** There is an update in agent metrics - trailing windows**
     + [Average handle time](metrics-definitions.md#average-handle-time): Build rules that run when the Average handle time historical metric reaches a specified duration.
     + [Agent occupancy](metrics-definitions.md#occupancy): Build rules that run when the Occupancy historical metric reaches a specified percent.
   + **There is an update in flow metrics - trailing windows of time**
     + [Flows started](metrics-definitions.md#flows-started): Build rules that run when the flow started count reaches a specified value.
     + [Flows outcome](metrics-definitions.md#flows-outcome): Build rules that run when the flow outcome count reaches a specified value for selected flow outcomes.
     + [Flows outcome percentage](metrics-definitions.md#flows-outcome-percentage): Build rules that run when the flow outcome percentage value reaches a specified percent for selected flow outcomes.
     + [Average flow time](metrics-definitions.md#average-flow-time): Build rules that run when the average flow duration reaches a specified duration for selected flow outcomes.
     + [Maximum flow time](metrics-definitions.md#maximum-flow-time): Build rules that run when the maximum flow duration reaches a specified duration for selected flow outcomes.
     + [Minimum flow time](metrics-definitions.md#minimum-flow-time): Build rules that run when the minimum flow duration reaches a specified duration for selected flow outcomes.

1. Choose **Next**.

## Step 2: Define rule actions
<a name="rule-actions-rtm2"></a>

1. Choose **Add action**. You can choose the following actions:
   + [Create Task](contact-lens-rules-create-task.md)
   + [Send email notification](contact-lens-rules-email.md)
   + [Generate an EventBridge event](contact-lens-rules-eventbridge-event.md): Use **Metrics Rules Matched** for the detail type.  
![\[The add action dropdown menu, a list of actions.\]](http://docs.aws.amazon.com/connect/latest/adminguide/images/contact-lens-add-action-no-wisdom.png)
**Note**  
 You can type @ to include the list of **agents, queues, flows or routing profile** that breached the metrics threshold within the **email** and **task** notifications. This list is automatically included within **resources** in EventBridge notifications.   

![\[The task action with variable injection.\]](http://docs.aws.amazon.com/connect/latest/adminguide/images/rules-rtm-task-action.png)


1. Choose **Next**.

1. Review and make any edits, then choose **Save**.