Triggers allow a deployed Zoho Zia Agent to act autonomously the moment a specific business event takes place, with no manual activation or custom code required. Whether a lead is created in CRM, a deal changes stage, or a support ticket is closed, a trigger can instruct the agent to respond immediately and independently.

How Triggers Work

A trigger connects two components:

• Publisher. A connection between Zia Agents and a Zoho service that monitors for events and reports them back to the agent platform. Publishers are configured at the organisation level in Settings and can be created inline during trigger setup.
• Event. The specific occurrence being monitored — such as a lead being created, a deal being updated, an invoice being approved, or a ticket being closed. The trigger rule defines: when this event happens on this publisher, run this agent, and pass it this data.

Prerequisites

• Agent deployment. The Triggers tab only appears after an agent has been deployed. Deployment ensures that tools have their parameter mappings and connections in place.
• Publisher availability. At least one publisher must exist for the target service, either pre-configured in Settings or created during the trigger setup process.

Configuration Steps

Step 1: Service and Publisher Selection

Select the Zoho service the trigger should monitor and choose an existing publisher or create a new one. New publishers can be set up without leaving the trigger configuration flow.

Step 2: Trigger Configuration

Four settings define the trigger's behaviour:

• Event Selection. Choose the specific event to monitor from a categorised list. Within Zoho CRM, for example, events are grouped by module (Leads, Accounts, Deals, and so on).
• Event Filtering. Decide whether the agent responds to every occurrence of the event (All Events) or only when the triggering record meets specific conditions. Multiple conditions can be combined for precise targeting.
• Session Key. Defines how trigger data is grouped before the agent processes it. The default is the record ID, which ensures each record's events are handled in context.
• Field Selection. Choose between passing all fields from the triggering record or selecting only the specific fields the agent needs. Selecting only relevant fields reduces token consumption, lowers cost, speeds up response, and limits data exposure.

Execution and Monitoring

When a trigger fires, it automatically initiates an agent run. These runs are visible in the Observability tab and are marked with "Triggers" as the Execution Source, making it straightforward to distinguish automatic runs from manual invocations or API-initiated calls.

Managing Triggers

From the Triggers tab you can enable or disable individual triggers using toggle switches, and edit or remove configurations through the action menu. An agent can be deployed to multiple targets, each with its own independently configured set of triggers.

Triggers vs. Integrations

Integrations give people a way to converse with the agent directly. Triggers allow events to instruct the agent to act, with no conversation required. This distinction matters when deciding whether your use case is conversational (use an integration) or event-driven (use a trigger).