Let's go over the steps necessary to implement click-to-call in a Python and Flask application.
- After the call connects, we provide TwiML instructions to connect the user to our sales or support teams.
Before we create our click-to-call application, we need to set up our environment first.
Let's put our Twilio credentials in a place where our application can access them. We'll store them as environment variables that our application can read.
# Twilio API credentials # Found at https://www.twilio.com/console export TWILIO_ACCOUNT_SID=ACXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX export TWILIO_AUTH_TOKEN=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX # You need to create a TwiML app to use this project. # Create one at https://www.twilio.com/console/phone-numbers/dev-tools/twiml-apps # then use its "Sid" value here export TWILIO_APP_SID=APXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX # Get your number at # https://www.twilio.com/console/phone-numbers/incoming export TWILIO_CALLER_ID=you_twilio_number
Twilio_Number placeholder values with your unique values, which you can find in the Twilio Console. You can use an existing Twilio phone number or obtain a new number.
For our solution, we'll need a form where the user can enter a phone number.
No need to overthink this step as the real goal is to
POST the user's phone number to your controller.
What information does this form need?
- An input for the User's phone number
- An input for the sales team phone number
Since the page doesn't need to render new content after clicking on submit, we decided to implement the POST action via Ajax using jQuery. Let's look at that code next.
Now that we have the front end done let's build the back end that will receive this data. We'll start our exploration in the next step.
First, we instantiate a
TwilioRestClient object with our Account SID and Auth Token. This is essentially our Python REST API handler, which we could use to send SMSes.
But for now, we need it to get access to the object that we're going to use to create phone calls.
Let's look at how to make a phone call.
Next, we'll use the
create method from the
calls object to make an outgoing phone call. This requires us to pass a From number, a To number and a URL parameter that tells Twilio what to do after it connects the call to our user. In this case, Twilio needs to dial in the Agent once the call has been placed. We'll discuss this more in the future steps.
Next, let's take a look at how to define the outbound endpoint.
Twilio makes a request to our application when the call is created using the REST API. We'll need to create an endpoint that is publicly available for internet requests: we'll walk through a way to make your localhost accessible via ngrok a little later in this tutorial.
Next, we will see how to generate the TwiML required for Twilio to take actions on behalf of our application.
TwiML is a set of simple verbs and nouns written in XML that Twilio reads as instructions.
In this case, our instructions inform Twilio to simply SAY something to the user. If we wanted to connect the user to a real support agent, we would have to add a Dial verb that specifies the support agent number so the customer can talk to him or her.
To make writing TwiML easy, many of the helper libraries have methods that generate TwiML for you. In this case, we use the
twilio Python library to create a TwiML response that will instruct Twilio to say something.
And with that, you've helped us get a working click-to-call form, ready to be integrated into your own application.
Now you can run and test your Twilio app.
However, you probably want to test it using a publicly available endpoint without having to go "public" with your app. The best option is to use ngrok.
Note: For more information about running the application, see the Readme file in the app github repository.
ngrok generates a secure URL that forwards traffic to a port, usually 5000, on your localhost server. It allows you to run applications locally but make them a publicly available endpoint via secure tunneling. This allows you to test your application and do all the things you want your app to do but in a secure public space.
In case you are wondering, ngrok is an executable (
ngrok.exe) that you run on the command line or terminal.
ngrok is brilliantly simple to use. You sign up on the ngrok website (ngrok.com), download the ngrok.zip file for your OS of choice, then unzip the file into an easily accessible location.
For example, on Windows, you can place it in your
\Users directory in
\AppData\Roaming\<ngrok-directory-name>\ngrok.exe. Then, update your PATH environment variables to the location of the ngrok executable. On Linux or OSX, it's much easier (see the instructions on the ngrok site).
And that's it.
So, now you are ready to use ngrok.
Before you start ngrok, it might be a good idea to have your Twilio Console open.
For Mac and Linux, open a terminal and run this command to start ngrok:
$ ./ngrok http 4040
On Windows, open a command prompt and run this command to start ngrok:
$ Path-to-ngrok> ngrok http 4040
Alternately, you can start ngrok using the following command:
$ ngrok http 4040 -host-header="localhost:4040"
The port number "4040" is arbitrary. If your local server is running on another port, replace "4040" in the command with the appropriate port number.
This will start ngrok. A running instance of ngrok will appear in the terminal showing the local web interface (in this case http://127.0.0.1:4040) and the public URL (in this case, https://28dd499b.ngrok.io).
The forwarding public URL needs to be set in the Twilio Console. In the Console, select the active phone number you set in the configuration file and enter the public URL from ngrok (shown above) in the "A Call Comes In" field under Voice & Fax. Be sure to select Webhook from the dropdown list.
Remember, each time you run ngrok, a new public URL will be generated. You will need to set this new URL in the Console.
And that's it. You now have a publicly available endpoint for your app, so now you can take a look at your running app and enter the phone numbers using your ngrok address: https://xxxxxx.ngrok.io
- Check out how iAdvize uses Twilo click-to-call to connect to online shoppers with customer support representatives
- Appointment Reminders - automate the process of reaching out to your customers in advance of an upcoming appointment.
- Automated Survey - instantly collect structured data from your customers with a survey conducted over a voice call or SMS text messages.
Thanks for checking out this tutorial! Tweet to us @twilio and let us know what you're building!