telecrm AI

Now that you have learned about creating a Chat IQ agent, let's build the complete workflow in this video to get it started, test it with practical examples, and see how the conversation moves between Chat IQ and the team.

Let’s take an example of a coaching institute that receives frequent WhatsApp enquiries about fees, batches, and demo classes. Let’s set up Chat IQ to handle routine questions and involve counsellors when needed.

Before You Build: Understand the Questions 

1. What triggers the workflow? A new incoming WhatsApp message.
2. What should Chat IQ handle? Routine questions about courses, fees, and batches.
3. When should a person take over? When a parent requests a discount or needs human assistance.

Once these things are clear, building the workflow becomes much easier. Let’s set it up step by step.

Part 1: Build the Workflow

• Navigate to the Workflows section on the web app. Click on the Create new option.
• Use a descriptive name so your team can easily identify its purpose.
• Next, select the Event: On WhatsApp Received.
• This ensures that the workflow starts whenever a parent sends a new WhatsApp message.
• The agent will now handle the remaining enquiries using its configured instructions and knowledge base.
• Chat IQ provides three possible exits, and each one serves a different purpose.
• Green — Success: Chat IQ has generated a reply.
• Red — Failure: Chat IQ could not complete the action successfully.
• Yellow — Handoff: Chat IQ has determined that a human needs to take over.
• We’ll configure all three paths so that every outcome has a clear next step.
• Let’s start with the green exit.
• Add a worker node and configure it to save relevant details from the conversation. Select the AI Agent variable so that the agent picks up all relevant information and answers from the Knowledge base saved. 
• This helps your team maintain useful lead information without updating everything manually.
• Next, Chat IQ failure follow-up. Connect the red exit to an action that alerts the admissions team.
• This ensures the team knows when Chat IQ encounters a failure and can follow up instead of leaving the parent’s enquiry unanswered.
• Finally, configure the yellow exit for human handoff.
• Assign the lead to a counsellor and create a follow-up task so the counsellor knows that their intervention is required.
• Once all three exits are configured, review the connections and save your workflow.
• Your basic automation is now ready for testing.

Part 2: Test the Workflow Like a Real Parent

Test 1: A Routine Course Enquiry

• First, turn on the workflow and send a WhatsApp message from your test number: ‘What are the fees for the Class 10 foundation batch?’
• Chat IQ should respond using the information available in its configuration.
• You should also see the corresponding action on the lead record.

Test 3: A Discount Request

• Now, test the handoff by sending: ‘Can you give me a 20 percent discount if I enroll today?
• If the agent’s instructions are configured to hand off discount requests, Chat IQ should trigger the yellow exit.
• The workflow should then assign the lead to a counsellor and create a follow-up task.

Test 4: Hand the Conversation Back to Chat IQ

• “Once the counsellor has handled the enquiry, reply manually and then use the Handback option to return the conversation to Chat IQ.
• Send another question to confirm that the agent can resume handling the conversation.
• You can also check the conversation tags to distinguish AI-generated replies from messages sent by a team member.

This is how the workflow manages different types of enquiries without requiring your team to handle every message manually.

And that’s all you need to implement! You’ve successfully built and tested a Chat IQ workflow. 

Start with one use case, monitor the results, and expand your automation as you become more comfortable with it.