AI Automation Builder

The AI Automation Builder allows specialists to create automation rules using natural language descriptions instead of manually configuring triggers, conditions, and actions. Sera AI interprets the description and fills in a complete automation rule that the specialist can review before saving. Specialists do not need to be experts in script or code writing.

Creating an automation rule with AI

1. Navigate to the Automation Rules section.
2. Click to create a new automation rule. The Sera AI prompt appears on a new rule before it has a name, when Sera AI is enabled for the account.
3. Describe what you want the rule to do in plain language. For example: "When a user says they can't send emails, classify the request as an incident."
4. Click Ask Sera AI. With that one natural language prompt, the automation is generated, including a trigger (for example, "On update of status"), conditions using pseudo code (for example, keyword matching or field value checks), actions to execute when conditions are met, and a description explaining what the rule does.

Review before saving

Sera AI fills in the rule editor; it does not save the rule. A rule is active as soon as it is saved unless the Disabled checkbox is ticked. To review a generated rule safely, tick Disabled before saving, check the trigger, conditions, and actions (small tweaks to the pseudo code are sometimes needed), and untick it when the rule does what you intended.

This review step ensures that AI-generated rules do not take unintended actions.

Generating a rule description

For a rule that already has a trigger and at least one action, Generate with Sera AI writes the rule's description from its logic. The suggestion can be accepted, retried, or cancelled.

Practical examples

• Inbox management: "Complete low impact requests assigned to David Wood" generates a rule that triggers on status update, checks for low impact and the assigned person, and completes matching requests automatically.
• Email classification: "When a user says they can't send emails, classify the request as an incident" generates a rule that classifies matching requests, applies the right SLA, and routes directly to the right team.
• Kanban routing: "Anytime a request comes in for the content creation service, add it to the to-do column in my Agile board" generates a rule that automatically places new requests on a specific board.

The value extends beyond IT. Teams across the organization that use Xurrent for service management (marketing, facilities, HR) can build automations for scenarios they could not handle before, without needing technical expertise.

Best practices

• Be specific in your description. More detail produces more accurate rules.
• Review the generated conditions carefully, particularly for edge cases.
• Save new rules as disabled, and test them on a limited scope before enabling them broadly.

‍