How to integrate Dialogflow Virtual Agent with Con...

avani28

Tired of the same old contact center solutions that struggle to keep up with the demands of today's customers? Imagine a contact center where intelligent virtual agents handle routine tasks effortlessly, allowing your human agents to focus on complex and meaningful interactions. Picture a world where customer service is available 24/7, delivering personalized assistance and resolving issues efficiently.

Google Cloud's Contact Center AI (CCAI) Platform, coupled with your own custom-built Dialogflow virtual agent, is the key to turning this vision into reality. This powerful combination brings the flexibility and conversational intelligence of Dialogflow together with the robust infrastructure and features of CCAI Platform, creating a truly transformative customer service experience.

In this step-by-step guide, we'll show you how to seamlessly integrate your Dialogflow virtual agent with CCAI Platform, unlocking a world of possibilities for your contact center. You'll learn how to leverage AI to automate routine tasks, improve customer satisfaction, and streamline your operations. Whether you're new to virtual agents or looking to optimize your existing setup, this guide will equip you with the knowledge and tools to take your customer service to the next level.

Why bring your Dialogflow Virtual Agent to CCAI Platform?

Elevate self-service: Free your human agents from routine inquiries by empowering your virtual agent to handle them with ease.
Always-on support: Ensure your customers receive immediate assistance 24/7, boosting satisfaction and reducing response times.
Multimodal, omnichannel customer experience: Deliver a unified brand experience across every touchpoint – voice and chat.
Scalability for growth: Effortlessly adapt to increasing customer interactions as your business expands.
Data-driven optimization: Leverage CCAI analytics to gain valuable insights into customer behavior and preferences, refining your virtual agent for continuous improvement.

Prerequisites

Before we dive into the integration, let's make sure you have the essentials in place:

Google Cloud Project: While not a strict requirement, keeping your Dialogflow CX agent and CCAI Platform instance within the same Google Cloud project could streamline the integration process and ensure seamless communication between the two services.
CCAI Platform Instance: You'll need an active CCAI Platform instance up and running.
Permissions: Verify that you have the Contact Center AI Administrator role within your Google Cloud project. This role grants you the necessary permissions to manage both Dialogflow and CCAI Platform resources.
Enable Dialogflow CX API: In the Google Cloud Console, navigate to the APIs & Services and enable the Dialogflow CX API. This step is crucial for allowing CCAI Platform to communicate with your virtual agent.
Create a service account: Generate a service account with the Dialogflow API Admin role within your Google Cloud project. This service account acts as a secure bridge between CCAI Platform and Dialogflow. Download the JSON key file associated with this service account; you'll need it in the next steps.

Step-by-step integration

Here's a step-by-step walkthrough of the process:

Step 1: Creating the Dialogflow agent

Before we dive into the CCAI Platform, we need to build the brains behind your virtual agent: the Dialogflow CX agent. This agent will be responsible for understanding customer inquiries, processing information, and crafting intelligent responses. Let's start by setting up your agent in the Dialogflow CX console.

Head over to Dialogflow CX and select your Google Cloud Platform (GCP) project.
Click on Create Agent and then choose Build your own.
Give your agent a descriptive display name (e.g., "flight_booking_agent").
Selecting the appropriate location and default source is important for ensuring optimal performance and functionality. For this example, choose global for the location and CX flow as the default source.

Image 1.png

Finally, click Create to bring your agent to life.

Image 2.png

Once your agent is created, you'll be taken to the default start page. This is where you'll begin designing and building the conversational flows that will guide your virtual agent's interactions with customers. Remember, a well-structured and comprehensive Dialogflow agent is the foundation for a successful integration with CCAI Platform.

Step 2: Configure the conversational flow

Now that you have your Dialogflow CX agent created, it's time to design the conversational flow. To streamline this process, you can import a pre-built agent that's already been designed to handle flight bookings. This will give you a solid starting point for customization, allowing you to define the paths a conversation can take based on user input and how the agent should respond. To get started with the pre-built agent:

Download the sample virtual agent.
Select View all agents in the Agent dropdown menu at the top of the Dialogflow CX UI.
Locate your agent (e.g., "flight_booking_agent") and click on the three dots in front of its name.
Select Restore from the expanded menu options.
Select the Upload radio button.
Click on select file and select the file we downloaded earlier then, click Restore.

Your agent should now look like this:

Image 3.png

Step 3: Test the Virtual Agent

Before connecting your virtual agent to CCAI Platform, it's crucial to thoroughly test the conversational flow within Dialogflow CX. This helps ensure that your agent understands user intents, extracts relevant information correctly, and provides accurate responses. To test the agent flow:

The Dialogflow CX console provides a built-in simulator where you can simulate conversations with your virtual agent.
Click on Test Agent, type in user queries and observe how the agent responds. Pay attention to whether it understands the intent correctly and provides the expected answers.

Image 4.png

By thoroughly testing your Dialogflow CX agent, you'll be well-prepared to integrate it with CCAI Platform, confident that it can handle real-world customer interactions effectively and deliver a seamless support experience.

Step 4: Configure the Agent Settings

Before integrating your Dialogflow agent with CCAI Platform, it's important to configure a few settings within Dialogflow itself to ensure optimal performance and debugging capabilities.

Navigate to the Agent Settings (gear icon) within your Dialogflow CX agent interface.
On the General tab, scroll down to the Logging settings section.
Ensure that logging settings are enabled. This will allow you to capture detailed logs of interactions with your virtual agent.

Image 5.png

By enabling logging, you'll gain valuable insights into how your virtual agent is performing, identify potential errors, and gather data to continuously improve its responses and capabilities.

Step 5: Configure the Agent Assist conversation profile

After designing and testing your Dialogflow CX agent's conversational flow, it's time to enhance its capabilities by integrating it with Agent Assist.

Agent Assist is a powerful tool that leverages Google's AI to provide real-time suggestions, summaries, and knowledge articles to your human agents during customer interactions. This integration can significantly improve agent productivity and enhance the overall customer experience.

Navigate to the Agent Assist page.
Click on Conversation profiles in the left sidebar menu and select your GCP project.
Click on Create. In the menu that appears, enter a unique and descriptive name for your conversation profile in the Display name box (e.g., "Flight Booking Conversation Profile").
Scroll down and toggle on Enable virtual agent and select your "flight_booking_agent" agent from the dropdown list. This connects Agent Assist to your Dialogflow agent.
You can explore and enable other features like Smart Reply or Article Suggestion based on your specific needs and use cases.
Click Create to save your conversation profile.

Image 6.png

By configuring a comprehensive Agent Assist conversation profile, you equip your live agents with the tools they need to provide efficient and accurate support. The AI-powered insights and suggestions will help them resolve customer issues faster, leading to improved customer satisfaction and overall contact center performance.

Step 6: Connect Dialogflow CX to CCAI Platform

Now that Dialogflow CX agent is set up and tested, it's time to establish a connection between Dialogflow CX and CCAI Platform. This integration allows CCAI Platform to utilize your virtual agent for customer interactions.

Sign in to your CCAI Platform instance.
Navigate to Settings and then click on Developer Settings.
Scroll down to the Virtual Agent Platform section and click on Add Platform.
Enter a descriptive name for the platform (e.g., "Flight Booking Virtual Agent").
Select Dialogflow CX as the service and click Create.
Upload the JSON service account key file that you downloaded from your Dialogflow project and click Save. (This service account should have the "Dialogflow API Admin" permission to ensure smooth communication between CCAI Platform and Dialogflow.)
After creating the platform, locate it in the Virtual Agent Platform section. Toggle the switch to the On position to activate the integration.

Image 7.png

By completing these steps, you've successfully established a connection between your Dialogflow CX agent and CCAI Platform. You're now ready to incorporate this virtual agent into your contact center workflows and enhance your customer interactions with intelligent automation.

Step 7: Configure the Virtual Agent in CCAI platform

Now that Dialogflow CX agent is connected to CCAI Platform, it's time to configure your virtual agent within CCAI Platform. This is where you'll specify how your virtual agent interacts with customers, which channels it supports (voice or chat), and how it integrates with your live agents.

Types of virtual agents:

CCAI Platform offers two types of virtual agents:

Customer Support: These are the virtual agents you're likely most familiar with. They handle customer inquiries, provide support, and can escalate conversations to live agents when needed. This is the type we'll focus on in this guide.
Task Assistant: Task assistants are designed to automate repetitive tasks for your human agents. They can handle things like updating customer records, filling out forms, or scheduling appointments, freeing up your agents to focus on more complex interactions.

Creating a customer support virtual agent

In the CCAI Platform interface, navigate to Settings and then click on Virtual Agent.
Click Add Virtual Agent and select Customer Support. This will open a new dialog for configuring your virtual agent.
Give your virtual agent a descriptive name (e.g., "Flight Booking Assistant").
Select your Dialogflow CX virtual agent platform (e.g., Flight Booking Virtual Agent) you created in the above step from the dropdown menu. Then, choose the conversation profile (e.g., Flight Booking Conversation Profile) you created in step 5.
Choose the channel(s) you want your virtual agent to support (e.g., "Voice").
Click Save, your customer support virtual agent is now configured and ready to start interacting with your customers.

Image 8.png

Click Back to Virtual Agents and toggle on.

Image 9.png

Now, your newly created customer support virtual agent will be listed in the Virtual Agents section of CCAI Platform. You can now integrate this virtual agent into your call flows, allowing it to interact with customers and handle inquiries based on the conversational flow you designed in Dialogflow CX.

Step 8: Configure queue settings

After configuring your virtual agent, the next step is to establish the queue within CCAI Platform to route calls correctly and connect them with your virtual agent.

Navigate to Settings from the main menu and then Queue and click on Edit/View IVR.
Add Queue name in New Item and click Done, this queue will associate with your virtual agent.

By completing these steps, you ensure that when customers call your Global Outbound Phone Number, they will be routed directly to the appropriate queue. However, to have your virtual agent handle these calls, you need to explicitly assign it to the queue.

Step 8.1: Assign virtual agent to queue

Navigate to the Queue you configured earlier

In queue specific settings, click on Assign Virtual Agent.
From the dropdown menu, select your virtual agent (e.g., "Flight Booking Assistant") and click Save. This will make your virtual agent the primary handler of calls routed to this queue.
Click Back to Settings and toggle on the Virtual Agent.

Now your virtual agent is connected to your specified queue, and it will start answering calls automatically based on the selected hours of operation.

Step 8.2: Transfer to a human agent

While your virtual agent is designed to handle a variety of customer inquiries, some situations may require the expertise of a live agent. CCAI Platform allows seamless routing from the virtual agent to a human agent to ensure a smooth customer experience. To enable this, you need to configure a human agent for the queue and specific route within your Dialogflow CX agent and include a custom payload that triggers the transfer to a human agent.

Navigate to the Queue you configured earlier
In queue specific settings, click on Assign Human Agents.
Add an agent to the cascade group and click Save.
Click Back to Settings and toggle on the Human Agents.

Step 9: Add an escalation route in Dialogflow CX

Go to your Dialogflow CX agent.
Click on Start Page, add new Route.
Enter description as “Representative”. Add a new intent named “call_agent” and add training phrases like these: talk to agent, human agent, agent, call agent and click Save.

Image 11.png

Enter “Let me route the call to agent” in the Agent Says field.
Click on Add dialogue option, choose Custom payload and paste the given payload and click Save.

Custom Payload:

{
"ujet": {
"type": "action",
"escalation_reason": "by_customer",
"action": "escalation"
}
}

Image 12.png

By implementing this transfer mechanism, you empower your virtual agent to seamlessly route complex or sensitive inquiries to human agents. This ensures that customers receive the most appropriate level of support, leading to improved customer satisfaction and problem resolution.

Step 10: Test thoroughly

After configuring your virtual agent in CCAI Platform, it's crucial to test it thoroughly. This ensures your virtual agent interacts smoothly with customers across all channels and that the integration with CCAI Platform works perfectly.

Dial the phone number linked to your CCAI Platform instance and have conversations that follow the expected flow of your Dialogflow CX agent. Confirm that the agent understands your intent, provides accurate responses, and seamlessly escalates to a live agent when needed.

To find the number your customers will dial:

Navigate to Settings then Call within your CCAI Platform interface.
Copy the Global Outbound Phone Number. This is the number your customers will dial to reach your contact center.

Take your customer service to the next level

As you've seen, integrating your Dialogflow virtual agent with CCAI Platform opens up a world of possibilities for transforming your customer service. By leveraging the strengths of both platforms, you can create a seamless, intelligent, and engaging experience for your customers while streamlining your operations and empowering your agents. The key to success lies in ongoing optimization and a commitment to delivering exceptional service. Now that you have the tools and knowledge, take the next step and start building a contact center that truly delights your customers.

Have questions? Ask me in the comments below!