Skip to main content

WhatsApp LLM Chatbot

Yeshawanth Reddy
Updated

You can now use WhatsApp as a communication channel for your LLM Chatbot alongside SMS and Email. This integration allows you to connect directly with candidates and clients on one of the world's most widely used messaging platforms.

Benefits of WhatsApp Integration

Using WhatsApp with your LLM Chatbot provides several advantages:

  • Increased engagement: Reach candidates and clients through a familiar and trusted messaging platform.
  • Flexible campaign support: Run both outbound chatbot campaigns and Text-to-Apply (T2A) experiences using WhatsApp.
  • Real-time communication: Automate conversations while delivering a seamless messaging experience.

Common Use Cases

You can use WhatsApp LLM Chatbots to:

  • Send job updates and interview confirmations.
  • Engage candidates and clients through automated chatbot conversations.
  • Launch Text-to-Apply campaigns using a dedicated WhatsApp number.
  • Include WhatsApp chatbot interactions as part of Workflow automations.

Before You Begin

Before using WhatsApp with your LLM Chatbot, the following prerequisites must be completed:

  • Business profile registration
  • Sender onboarding
  • WhatsApp template approval
  • WABA chatbot number assignment

Note: Contact your Sense Customer Success Manager (CSM) or the Product Operations team to complete these prerequisites.

Configuring WhatsApp for LLM Chatbots

Enable WhatsApp for a Flow

A new WhatsApp toggle is available in Flow Settings.

When enabled, the selected chatbot flow becomes available through WhatsApp, allowing you to choose which chatbot experiences should be delivered using this channel.

Configure Outbound WhatsApp Bots

When WhatsApp is enabled for an outbound chatbot, you can configure the message candidates receive before the conversation begins.

Available options include:

  • View the default WhatsApp greeting template.
  • Browse your agency's approved WhatsApp templates.

  • Select a different approved template if required.
  • Configure template variables and free-text placeholders such as:
    • Candidate name
    • Job title
    • Other supported dynamic fields

Configure Text-to-Apply (T2A) Bots

Text-to-Apply bots can now be launched through WhatsApp.

With this configuration:

  • Candidates start the conversation by sending a predefined keyword to a WhatsApp number.

  • You can configure a custom greeting message since the conversation is initiated by the candidate.
  • Chatbot WhatsApp numbers and short codes can be assigned directly from Agent Channel Configuration.

Test the WhatsApp LLM Chatbot

After configuring your WhatsApp LLM Chatbot, you can test the LLM Chatbot directly within Agent Builder to verify the end-to-end flow and overall user experience before making it available to candidates.

Before You Begin

Before testing the chatbot, ensure that:

  • WhatsApp is enabled for the flow.
  • A WhatsApp chatbot number has been assigned to the flow.
  • The required WhatsApp template is configured for outbound chatbot flows.
  • All WhatsApp onboarding prerequisites have been completed.
  • The chatbot has been published.

Configure the Test

  1. Open the required agent in Agent Builder.
  2. Navigate to the Test tab.
  3. From the Channels panel, select Chat (WhatsApp).
  4. Click Testing Configuration.
  5. Configure the following settings:
    • Job Title – Select the job to use during the test.
    • Choose Candidate – Select the candidate whose information will be used during the conversation.
    • Enable write back during test (Optional) – Enable this option if you want the chatbot to write test interactions back to the selected candidate record.
  6. Click Save to apply the test configuration.

Test the Conversation

After saving the test configuration, the WhatsApp conversation preview is displayed within the testing window.

You can:

  • Verify the initial greeting message displayed to the candidate.
  • Simulate a real WhatsApp conversation by sending test messages.
  • Validate the chatbot responses and conversation flow.
  • Verify that configured variables and WhatsApp templates are rendered correctly during the interaction.

Note: The in-app testing experience simulates the WhatsApp conversation within Agent Builder and allows you to validate the chatbot configuration before deploying it to candidates. This helps ensure the chatbot behaves as expected and provides the intended conversational experience.

WhatsApp Integration with Workflows

WhatsApp support is also available within Workflows, making it easier to automate chatbot conversations.

WhatsApp Chatbot Node

A dedicated WhatsApp Chatbot node is available in the Workflow Builder.

Using this node, you can:

  • Select from published WhatsApp LLM Chatbots.
  • Send WhatsApp chatbot conversations as part of an automated Workflow.
  • Incorporate WhatsApp interactions into existing recruitment automation processes.

WhatsApp Message Types

WhatsApp supports two different message types depending on how the conversation is initiated.

Templated Messages

Templated messages are pre-approved WhatsApp templates used to start or restart conversations.

Characteristics

  • Require WhatsApp approval before use.
  • Support dynamic placeholders such as candidate name or job title.
  • Used to initiate new conversations or re-engage candidates outside an active session.

Session Messages

Session messages are sent during an active WhatsApp conversation.

Characteristics

  • Can be sent within the 24-hour messaging window after a candidate sends the first message.
  • Do not require prior WhatsApp approval.
  • Allow more flexible and conversational interactions between the chatbot and the candidate.

Outbound Message Preview

To provide greater visibility into the candidate experience, the outbound bot configuration now includes an If sending via WhatsApp section.

This preview displays exactly how the selected WhatsApp message template will appear to candidates before it is sent.

Recommendations

To get the most value from WhatsApp LLM Chatbot integration, we recommend the following:

  • Enable WhatsApp for your outbound chatbot campaigns.
  • Create Text-to-Apply chatbot experiences using a dedicated WhatsApp number.
  • Test your approved WhatsApp templates before launching campaigns.
  • Share any implementation feedback or issues with the Product or QA teams to help improve the experience.