Creating a Pipeline

About this task

A pipeline dictates how incoming data feeds are evaluated and where the resulting events are sent. The creation process is divided into four primary configuration nodes.

Procedure

  1. Navigate to the Pipelines section to start a new pipeline.
    1. In the left navigation pane, click Unica Detect and select Pipelines.
    2. Click + New Pipeline.
    3. In the Create Pipeline dialog box, enter a Pipeline Name (e.g., recharge_Customer) and a Pipeline Description.
    4. Click Next. The Pipeline visual configuration canvas appears.
  2. Configure Step 1: Who is this Pipeline for? (Audience)
    1. On the first node, click + Set Audience.
    2. In the configuration pane, click + Add Raw Condition.
    3. Select an attribute from the drop-down menu (e.g., country).
    4. Set the operator (e.g., Not Equal To) and enter the corresponding value (e.g., "" to target all available audiences).
  3. Configure Step 2: What will start the Pipeline?
    1. On the second node, click + Set Actions.
    2. From the Select Feed drop-down menu, choose your pre-configured feed (e.g., Recharge).
    3. Under the Trigger Matching section, define the condition that must be met to trigger the event. Select a Tuple Attribute (e.g., TransactionAmount), set the operator (e.g., Greater Than), and enter the threshold value (e.g., 200).
  4. Configure Step 3: Where to send this data? (Destination)
    1. On the third node, click + Set Data Destination.
    2. In the Event ID field, enter a unique identifier for the event (e.g., rechargeCust).
    3. From the Destination drop-down menu, select your downstream system (e.g., MariaDB).
    4. In the Attributes section, select the checkboxes for all data points (e.g., AUDIENCE_KEY, transactionAmount, you want to forward to the destination from tupleAttributes and profileAttributes.
      • Profile Attributes: Customer-level data stored in the customer profile (e.g., firstName, city, gender, emailID). These are relatively static and represent who the customer is.
      • Tuple(Transactional) Attributes: Event-level data fields that come from incoming real-time data feeds (transactions/events), and its attributes are the individual fields within that record — for example, transactionAmount, eventTime, duration, etc.
    5. Select the Mark all events as non-marketing checkbox to disable the default global contact policy check. This default is currently set to limit the outgoing events to 50 in 7 day, per customer, across all events. There is also an event level contact policy that can be configured in the When Should it run step.
    6. Click Save.
  5. Configure Step 4: When should it run? (Schedule and Contact Policy)
    1. On the fourth node, click + View Schedule.
    2. Select the active days for the pipeline (e.g., check the boxes for Monday through Sunday).
    3. Use the slider or time fields to define the active hours for each day (e.g., 0:00 to 23:59). You can configure one day and click Copy timing to to apply it to other days.
    4. Click Apply.
    5. In the left configuration menu, select Contact Policy.
    6. Set the Total Triggers Limit and Per User Limit (e.g., maximum 100 times within 1 day).