How to set up the Dialogflow chatbot template step by step

This is one of the reasons why we have started to expand our catalog of templates, in which you will find at least one that can help you.

Among them is our Dialogflow template, which combines the potential of Artificial Intelligence developed by Google with our multichannel platform.

We tell you how to start making the most of this template. 

Template features

This is one of our ready-to-use templates, although in this case it is necessary to perform a series of steps to integrate Dialogflow that we will explain in detail below.

It is configured to work on the following channels:

• WhatsApp
• Facebook
• Instagram
• Webchat

Configure Dialogflow for integration

First we go to Dialogflow to start the configuration with the selected Google account.

1. We choose the “Create Agent” option, starting the process.
2. We name the agent and click on the “Create” option.

3. The newly created agent will appear in the left column. To proceed with the integration, click on the nut to access the settings and go to the link that presents the “Project ID” section.

4. The objective now is to create a Service Account. So, in the left column we select “IAM and administration” and in the options presented, we choose “Service Accounts”.

5. Select the “Create service account” option for our integration.
6. We fill in the account details and select “Create and continue” when all the data is complete.

7. In the permissions section, we select the “Dialogflow API Administrator” role and click “Continue” when we are done.

8. At this point, we want to create a Key for our account. First we click on the link that appears in the Email section.

9. In the top menu go to “Keys” > “Add Key” > “Create new key”.

10. The type of key we will use is a JSON file. Select this option and create it.
11. The file will be downloaded to our computer.

Installing Dialogflow integration in WOZTELL

1. We log in with our account and select the “Integrations” option in the top menu and navigate to select a “New Integration”.
2. Search for “Dialogflow” and confirm your selection by clicking on the “Get” button.

3. We make sure to select the latest version and confirm again by clicking on “Get”.
4. We enter a name for the integration in the “Alias” field. This field is mandatory, while the description is completely optional. Once the required information has been filled in, select “Install”.

Configuring the Dialogflow integration

1. In the same “Integrations” section we now select the one we have just installed.
2. To complete the integration, we will need to enter a series of data. The information is contained in the json file that we have previously downloaded.

It is very important that when copying the Private Key we separate the lines correctly, adding the line breaks. Basically, we have to search in our text for the “\n” characters (line break encoding) and replace them with a line break. To locate the separations, we recommend using the “CTRL+F” search engine and searching for the “\n” characters. This combination must be removed so that the text looks similar to the one shown in the image.

Configure the chatbot template

1. Import the template, and once opened in the Chatbot Builder, select the node called “DIALOGFLOW”.

2. We scroll until we reach the “Actions” section.
3. Now, we take care of selecting the integration that we have previously prepared. The information will appear automatically and in Locale we will select the language of our chatbot. Finally we click on “Save”.

4. If we scroll down to Tree Manager, the Config Map panel will open where we can configure whether we want the chatbot to restart after a period of time.

If you want it to restart, leave the value as “true”, in case you don’t want it to restart, write the value in “Restart Chatbot” as “false”.
In “Chatbot Restart Time” set the number of days/hours/minutes that must pass before the chatbot restarts. It is essential to configure “Time unite” with one of these values: “days”, “hours” or “minutes”.

Configure Live Chat

3. We move to “Action and parameters” to configure the conditions that the text fulfills for the chatbot to activate the live chat. For it to work correctly it is essential to configure this field as shown in the image.

4. Finally, in “Responses” we can configure a series of responses that the chatbot will show once it has identified the intention as wanting to talk to an agent.

Types of chatbot responses

In the case of wanting to configure a response as something other than text, it is necessary to configure it according to the channel to be used. Here we link the types of messages accepted according to the platform:

• WhatsApp
• Facebook
• Instagram
• Webchat

Since in this case the message needs to be configured at the code level, we recommend that a programmer takes care of it.

Conclusion

We hope this guide has helped you with the configuration of your new chatbot.

Artificial intelligence in chatbots is a great ally to alleviate the workload of agents and provide instant attention that will always make your customers feel heard.

This installation is just the beginning, now you just need to train and add more knowledge to your chatbot so that it can take care of the communications of all the channels you want.