Skip to content

Notifications Action

The Notifications action alerts people the moment the Flow runs, through In App, Email, Slack, Microsoft Teams, or PagerDuty. Every channel follows the same dispatch pipeline. The trigger event supplies the context, Qualytics resolves the message variables, and each channel receives a payload built for its format.

Writing the body

The message a notification carries, how to write and format it, and the tokens it can use live on the Message page.

Why Use a Notifications Action

Use a Notifications action when people should hear about a data quality event without watching the platform for it. Common cases:

  • Close the gap between detection and action: An alert lands where the team already works the moment a scan fails or an anomaly appears, instead of waiting in a dashboard.
  • Reach the right audience: Page on-call through PagerDuty, brief the team in Slack or Teams, keep stakeholders in the loop by Email, and surface events In App for the people already in Qualytics.
  • Carry the context along: Message variables put the datastore, container, and anomaly details in the message itself, so responders understand the issue before opening the platform.
  • Combine channels from one event: The same trigger can feed several notification actions as sibling branches, each with its own destination.

The Dispatch Pipeline

flowchart TD
    A[Flow Trigger Activates] --> B[Resolve Message Variables]
    B --> C{Notification Channel}
    C -->|In App| D[Platform Notification]
    C -->|Email| E[Email with Subject and Body]
    C -->|Slack| F[Interactive Message]
    C -->|Microsoft Teams| G[Adaptive Card]
    C -->|PagerDuty| H[Incident Event]

When a Flow trigger activates, Qualytics processes each notification action in order:

  1. Trigger event occurs: An anomaly is detected, an operation completes, an anomalous table or file is detected, or a scheduled or manual trigger activates.
  2. Variables are resolved: Dynamic tokens like {{ datastore_name }} and {{ anomaly_message }} are replaced with real values from the event context. The available tokens depend on the trigger type (see Message Variables).
  3. Payload is built: Qualytics constructs the appropriate payload for each channel: the Markdown you wrote is rendered for In App and Email, Slack builds an interactive message, Microsoft Teams an Adaptive Card, and PagerDuty a plain-text incident summary.
  4. Notification is dispatched: The message is sent to the configured destination. Each channel is processed independently, so if one fails, the others still deliver.
  5. Execution is recorded: Each notification records the message it sent and its result on the Flow's execution, linked to the event that triggered it, so you can trace what was sent and when in the execution history.

Tip

You can add multiple notification actions to a single Flow. For example, send an In App alert to the team, a Slack message to a monitoring channel, and a PagerDuty incident for critical anomalies, all triggered by the same event.

One notification per matching event

There is no cap on how many notifications a Flow can send. An event-based trigger runs once per matching event, so an Anomaly Detected Flow sends one message per anomaly that matches its filters. Narrow the trigger's filter conditions to keep the channel readable; see Actions Best Practices.

Channels

Pick the tab for your channel.

In App notifications deliver alerts inside the Qualytics platform. A new one pops up as a toast in the bottom-right corner of the screen and lands in the bell icon on the navigation bar, where each user views and manages their notifications. Who receives one depends on the event:

  • Users in a team associated with the event's target object (such as the datastore or container involved) receive it, whatever their role or team permission.
  • When the event has no target object (Schedule and Manual triggers), or the target object has no assigned teams, the users of every team receive it.
  • Admin users receive every In App notification, whether or not they are in those teams.

For step-by-step setup, see Add an In App Notification.

flowchart TD
    A[Flow Trigger Activates] --> B[Resolve Message Variables]
    B --> C[Identify Target Object Teams]
    C --> D[Send to Admins + Team Members]
    D --> E[Toast Appears and Lands in the Bell Icon]

For symptoms and fixes, see Actions Troubleshooting.

Email notifications deliver data quality alerts directly to one or more email inboxes. When a Flow trigger activates, Qualytics resolves any message variables, builds an email with the configured subject and body, and sends it to every address in the recipient list. Each address is processed independently. If one address is invalid, the remaining recipients still receive the notification.

Email is well suited for stakeholders who need a persistent, searchable record of data quality events outside of the Qualytics platform, such as compliance officers, external partners, or team members who do not regularly log in.

The email arrives formatted, in the Qualytics layout: a dark band carrying the logo, a single reading column, and your Markdown rendered as headings, lists, and links. Above the body sits a small label naming the kind of event, such as Anomaly detected or Operation completed. Qualytics writes that label from the trigger, so it is not something the message can set. When the trigger is about a resource, the default message closes with a link back to it in Qualytics; a scheduled or manual run has no such resource, so its default carries no link. Any link you write yourself renders the same way.

Every email also carries a plain-text copy of the same message. Clients that do not render HTML show that copy, and some inboxes use it to build the one-line preview beside the subject.

Info

Notification emails are sent through the Email integration when one is connected, or otherwise through the email service configured for your deployment, when it has one. If neither is available, email notifications are not delivered.

For step-by-step setup, see Add an Email Notification.

flowchart TD
    A[Flow Trigger Activates] --> B[Resolve Message Variables]
    B --> C[Build Email with Subject & Body]
    C --> D[Send to Configured Addresses]
    D --> E[Email Delivered to Inbox]

For symptoms and fixes, see Actions Troubleshooting.

Warning

Before using Slack notifications in Flows, you need to connect the Slack integration in Settings > Integrations. See the Slack Integration setup guide for instructions.

Slack notifications post data quality events into a channel of your Slack workspace. When a Flow trigger activates, Qualytics composes the Slack message for that event, with the event's details, a color for the result, and action buttons, and posts it to the configured channel. Responders can view details in Qualytics, acknowledge anomalies, add comments, or archive records without leaving Slack.

The formatted message is a template per event type. Every trigger type has its own layout, and so do the events an action produces, each carrying the details and controls that fit that event:

Event What the message carries
An operation completed The datastore, the operation type and its result, the message the operation recorded, and a View Operation button.
An anomalous table or file was detected The datastore, the scanned table or file, how many anomalies came out of it, and a View Results button.
An anomaly was detected The datastore, the table or file, the anomaly's type and message, a View Anomaly button, an Acknowledge button, and a menu holding Comment and Archive.
An anomaly changed status The anomaly's details, the status it moved from and to, and a View Anomaly button.
The Flow archived or deleted anomalies How many anomalies the action touched and the inactivity period it applied.
The Flow created or updated a ticket The ticket's number and status, the anomaly behind it, and a View Ticket button.
A scheduled or manual run The Flow's name, for a message you write yourself.

Acknowledge, Comment, and Archive act on the anomaly in Qualytics without leaving Slack, so your instance has to be reachable from the internet for them to work. The Preview section in the settings panel shows the template the Flow's trigger will use.

For step-by-step setup, see Add a Slack Notification.

flowchart TD
    A[Flow Trigger Activates] --> B[Resolve Message Variables]
    B --> C[Build Slack Message]
    C --> D[Send to Selected Channel]
    D --> E[Message Appears in Slack]
    E --> F{User Action}
    F --> G[View in Qualytics]
    F --> H[Acknowledge Anomaly]
    F --> I[Comment or Archive]

For symptoms and fixes, see Actions Troubleshooting.

Warning

Before using Microsoft Teams notifications in Flows, you need to connect the Microsoft Teams integration in Settings > Integrations. See the Microsoft Teams Integration setup guide for instructions.

Microsoft Teams notifications post data quality events into a channel of your Teams workspace as Adaptive Cards. When a Flow trigger activates, Qualytics resolves the message variables, builds the card for that event, and posts it to the selected channel. The card lays the event's context out in fields, with a color for the operation result, so the team reads it without leaving Teams.

Like Slack, the card is a template per event type. Every trigger type has its own card, and so do the events an action produces, such as a ticket created or updated or an anomaly archived or deleted by the Flow, each carrying the details that fit that event. The Preview section in the settings panel shows the template the Flow's trigger will use.

For step-by-step setup, see Add a Microsoft Teams Notification.

flowchart TD
    A[Flow Trigger Activates] --> B[Resolve Message Variables]
    B --> C[Build Adaptive Card]
    C --> D[Send to Selected Channel]
    D --> E[Card Appears in Teams]

For symptoms and fixes, see Actions Troubleshooting.

Warning

Before using PagerDuty in Flows, you need to connect the PagerDuty integration in Settings > Integrations. See the PagerDuty Integration setup guide for instructions.

PagerDuty notifications open an incident in your PagerDuty service for the event, so anomalies and operation results reach whoever is on call. The message you write becomes the incident summary, which PagerDuty also shows as the alert title, so your wording is the headline a responder reads. When the Flow runs, that summary is sent as you wrote it, with only its variables resolved. The severity you pick decides how PagerDuty routes and escalates it, following that service's own urgency rules.

For step-by-step setup, including severity levels, additional details, and routing key overrides, see Add a PagerDuty Notification.

flowchart TD
    A[Flow Trigger Activates] --> B[Resolve Message Variables]
    B --> C[Build Trigger Event Payload]
    C --> D[Apply Severity & Custom Details]
    D --> E[Route via Integration or Override Key]
    E --> F[PagerDuty Creates Incident]
    F --> G[On-Call Team Notified]

For symptoms and fixes, see Actions Troubleshooting.

Testing Notifications

Email, Slack, Microsoft Teams, and PagerDuty include a Test Notification button that sends a sample message before you publish the Flow, so you can verify the integration connection, the message formatting, and the channel targeting without waiting for a real event. In App has no test: the channel offers no test path, so the panel shows no button.

A test also runs the same check as publishing: a message or subject that names a variable the trigger does not provide is refused before anything is sent. The verdict appears under the button as one of three: Notification successfully sent, in green; a red alert titled with the channel, for example Email notification has invalid settings, listing each problem, when the template was refused; or Email notification could not be sent, with the channel's own response, when the settings passed and the delivery failed. A test that gets past this check, whether or not the channel then delivers, clears the red mark a refused action carries on the canvas. Tests fill variables with short sample values, so a passing test says the template is sound, not how long a real value will be.

Warning

Test behavior varies by channel. Email, Slack, and Microsoft Teams receive a sample message. For PagerDuty, a test creates an actual incident in your PagerDuty service.

See Also