Blog Post

Azure Communication Services Blog
5 MIN READ

Build a low code conversational SMS bot with Azure Communication Services and Azure OpenAI

ddematheu's avatar
ddematheu
Icon for Microsoft rankMicrosoft
Feb 22, 2023

In our last blog, we walked through the process of building an SMS conversational bot. Today we want to showcase a similar bot, but this time built completely on top of low code connectors for Power Automate. It is incredibly easy and exciting to leverage Azure Communication Services connectors in conjunction with Large Language Models (LLM) like ChatGPT to build a conversation bot. For this blog, we will be building Anton, a personal trainer bot for Contoso Gym. See the preview:

 

 

To follow along you will need:

 

This application will leverage Azure Event Grid to listen for incoming text messages to Azure Communication Services number, and an Azure Function to process the event and respond with an Azure OpenAI generated response.

 

We will start by configuring an Event Grid trigger for your Power Automate flow. This trigger will automatically connect to your existing Azure subscription and to the Azure Communication Services resource to configure your event subscription.

 

 

To keep the data organized, you will add a Parse JSON connector to help us parse the event. The event schema follows this structure:

 

Schema (click to see)
{
    "properties": {
        "data": {
            "properties": {
                "From": {
                    "type": "string"
                },
                "Message": {
                    "type": "string"
                },
                "MessageId": {
                    "type": "string"
                },
                "ReceivedTimestamp": {
                    "type": "string"
                },
                "To": {
                    "type": "string"
                }
            },
            "type": "object"
        },
        "dataVersion": {
            "type": "string"
        },
        "eventTime": {
            "type": "string"
        },
        "eventType": {
            "type": "string"
        },
        "id": {
            "type": "string"
        },
        "metadataVersion": {
            "type": "string"
        },
        "subject": {
            "type": "string"
        },
        "topic": {
            "type": "string"
        }
    },
    "type": "object"
}

 

Add that schema to the Parse JSON connector and add the event body as the content.

 

Next, we must add a call to Azure OpenAI to ask the model to generate a response. We will use REST APIs to POST a request with our prompt. For the prompt, we will use a combination of the message sent by the user and a pre-designed text. In this example, we want the GPT-3 model to act like Anton, a personal trainer at Contoso Gym. We added some sample quotes from which the model can draw inspiration. These quotes help guide the model’s response and provide a more intuitive and conversational flow for the users. See the completed prompt below:

 

 

You're Anton Stanescu and you are a personal trainer at Contoso Gym. You are having a conversation with one of your students. You are trying to motivate them to work out.
Examples of the types of things that Anton says:
- The only bad workout is the one that didn't happen.
- Strive for progress, not perfection.
- You won't get the muscle without the dedication and the drive
- Train hard...so you can stop a train if you need to.
- Push up, push down, push up, push down
- Exercise should be regarded as tribute to the heart.
Anton: Welcome to the Contoso Gym! Big things await for you.
Student:

 

 

To keep the flow organized, we will add a compose connector for our prompt so we can easily reference it later.

 

To communicate with the Azure OpenAI service, we will use POST request inside of an HTTP connector. In return, we will receive a text generated by the Azure OpenAI model. To configure the HTTP connector, you will need the URI to your Azure OpenAI resource. You will also need the `api-key` for your Azure OpenAI resource.

 

 

<base_url> + /openai/deployments/ + <deployment_name> + /completions?api-version=2022-12-01

 

 

 

Using the compose output for our prompt, we will concatenate together the prompt, the message the user sent through SMS, and a stop trigger for the Open AI model to stop generating text:

 

 

concat(outputs('Compose'), body('Parse_JSON')?['data']?['Message'], '\nAnton:')

 

 

Once we have the HTTP connector d, we will use a Parse JSON connector to process the response from Azure Open AI. The expected schema from Azure Open AI takes the following format:

 

Schema (click to see)
{
    "properties": {
        "choices": {
            "items": {
                "properties": {
                    "finish_reason": {
                        "type": "string"
                    },
                    "index": {
                        "type": "integer"
                    },
                    "logprobs": {},
                    "text": {
                        "type": "string"
                    }
                },
                "required": [
                    "text",
                    "index",
                    "logprobs",
                    "finish_reason"
                ],
                "type": "object"
            },
            "type": "array"
        },
        "created": {
            "type": "integer"
        },
        "id": {
            "type": "string"
        },
        "model": {
            "type": "string"
        },
        "object": {
            "type": "string"
        }
    },
    "type": "object"
}

 

We will configure it to use the body of the http request as the content the schema above.

 

Finally, we will configure our SMS connector for Azure Communication Services to respond with the new response generated by Azure OpenAI. You will need a connection string for your Azure Communication Services resource to initialize the SMS connector.

 

 

Now that everything is connected, test the flow by sending an SMS from your phone to the phone number you acquired through the Azure Communication Services resource.

Congratulations! using Power Automate. Azure Communication Services provides out of the box connectors for SMS, Email and Chat. Whether it is having Anton, the personal trainer bot talk to your customers or having a real person interact with them, Azure Communication Services can help. To learn more about other resources that Azure Communication Services offers see the links below:

Updated Feb 22, 2023
Version 1.0
  • davidlime160's avatar
    davidlime160
    Copper Contributor

    To build a low-code conversational SMS bot with Azure Communication Services and Azure OpenAI, we can follow these steps:

    1. Create an Azure Communication Services resource
    2. Create an Azure Function App to handle incoming SMS messages
    3. Use Azure OpenAI to generate responses to incoming SMS messages
    4. Use the Azure Communication Services SMS API to send responses back to the user

    Here are the details for each step:

    1. Create an Azure Communication Services resource:

      • Go to the Azure portal and create a new Communication Services resource.
      • Once the resource is created, note down the Connection String and Access Key as we will use these later.
    2. Create an Azure Function App to handle incoming SMS messages:

      • Go to the Azure portal and create a new Function App resource.
      • Once the resource is created, create a new Function within the app to handle incoming SMS messages.
      • Use the "HttpTrigger" template for the Function.
      • Configure the Function to listen for incoming SMS messages from the Azure Communication Services SMS API.
      • Use the Connection String and Access Key from the Communication Services resource to connect to the SMS API.
    3. Use Azure OpenAI to generate responses to incoming SMS messages:

      • Use the OpenAI API to generate responses to incoming SMS messages.
      • You can create an OpenAI API key and set up the OpenAI API integration in the Azure portal.
      • Use the OpenAI API to generate responses to incoming SMS messages within the Azure Function.
      • Store the generated responses in a database or a queue for later use.
    4. Use the Azure Communication Services SMS API to send responses back to the user:

      • Once the response is generated by Azure OpenAI, use the Azure Communication Services SMS API to send the response back to the user.
      • Use the Connection String and Access Key from the Communication Services resource to connect to the SMS API.
      • Use the SMS API to send the response back to the user's phone number.

    That's it! With these steps, you should now have a low-code conversational SMS bot up and running.

    Source: (external link removed by moderator)