Set up an outbound Voice AI agent call

Set up an outbound Voice AI agent call that proactively reaches out to customers. Use a Journey flow to initiate the call, pass customer context to the AI agent, and let the agent handle the conversation naturally. Common use cases include payment reminders, appointment confirmations, and proactive notifications.

Prerequisites [#prerequisites]

• Infobip account. If you do not have an account, sign up for an account.
• AgentOS enabled. Contact your dedicated account manager or Support if AgentOS is not available on your account.
• AI Agents Manager role assigned.
• Voice channel enabled. Contact your Infobip account manager to enable it.
• A phone number registered with Infobip for outbound calls.
• A published AI agent configured with voice-optimized instructions. If you have not created one, follow the Set up an inbound Voice AI agent tutorial first.

Products and channels [#products-channels]

• AI Agents
• Journeys
• Voice

Implementation [#implementation]

Step 1: Create the Journey flow [#step-1-create-the-journey-flow-implementation]

1. In the Infobip web interface, go to Automation Studio.
2. Create a new Journey.
3. Select an entry point for your outbound campaign:
   • Predefined audience: Use this to target a specific segment from your customer data.
   • Trigger: Use this to run the campaign at a specific time.
   • External source: Use this for existing primary journeys and third-party integrations or APIs.

Step 2: Pass context variables [#step-2-pass-context-variables-implementation]

1. Define flow variables that carry customer-specific data. For example, customer name, account number, appointment date, or payment amount.
2. Map these variables from your data source (Customer Profiles segment, API payload, or CSV upload) to the flow variables.
3. Reference these flow variables in the AI agent system prompt as context variables so the agent can personalize the conversation from the start.

Step 3: Add the Call element [#step-3-add-the-call-element-implementation]

1. Add a Call element to the flow to initiate the outbound call.
2. Select your outbound sender (phone number) in the From field.
3. In the To field, map the destination phone number from your flow variables or audience data.

Step 4: Add the Transfer call to AI agent element [#step-4-add-the-transfer-call-to-ai-agent-element-implementation]

1. In the Call answered branch of the Call element, add the Transfer call to AI agent element.
2. Select your published AI agent from the Choose AI agent list.
3. Select the Agent type: Speech to speech or Speech to text to speech.
4. Configure the voice settings for the selected type.

Step 5: Configure call session outcomes [#step-5-configure-call-session-outcomes-implementation]

Connect the exit branches of the Transfer call to AI agent element:

• Transfer to human agent requested: Connect to an Inbox queue or external contact center.
• AI agent session ended: Add a closing message and a Hang up element.
• Call not transferred: Add fallback handling or an error notification.
• Call ended: Log the event.

Step 6: Test and launch [#step-6-test-and-launch-implementation]

1. Validate the flow.
2. Run a test with a small audience or a single test number.
3. Verify that context variables are passed correctly and the agent personalizes the conversation.
4. Launch the flow for your full audience once testing is complete.

Next steps [#next-steps]

• Write voice-optimized system prompts, verify your setup, and test. See Prepare Voice AI agent for launch.
• Review the element configuration fields. See Transfer call to AI agent.
• Explore the AI Agents documentation.