Creating an Customer Support AI Assistant for Email Response: Step-by-Step Guide

In manual customer support, businesses often struggle with delays in responses and limited availability, which can lead to decreased customer satisfaction and loss of potential revenue. However, with AI in customer support, embodied in an AI-powered customer service system, these challenges are alleviated. AI-based customer support enables businesses to provide round-the-clock support, ensuring prompt responses to customer inquiries at any hour. This not only enhances efficiency but also boosts customer satisfaction levels. By leveraging AI customer services, companies can overcome the limitations of manual support and maintain a competitive edge in the market. Additionally, AI chat support facilitates personalized interactions, further enhancing the customer experience.

By following the detailed step-by-step instructions outlined below, you'll be able to set up your own automated email response system in no time. This system will enable your customer support chatbot AI Assistant to handle customer inquiries seamlessly, referencing attached files as needed to provide comprehensive answers.

Let's dive in and see how to create this powerful tool to enhance your ai customer support workflow.

Requirements:

• ChatGPT Assistant API access
• Latenode account

Step 1: Creating GPT Assistant & API Key

Log in and Navigate to Assistants:

• Log into your OpenAI account and navigate to the assistants section.

Create a New Assistant:

• Click on the "Create" button in the upper right corner.
• Provide a name for your assistant, such as "AI Customer Service Bot."
• Specify instructions for how you want it to respond, including context about using attached files.
• Choose a model and select available tools, enabling file retrieval.

Add Files:

• Upload any files that the assistant may need to reference, ensuring file retrieval is enabled.

Generate API Key:

• Copy the assistant's ID.
• Navigate to API keys and create a new key.
• Save the API key for later use.

Step 2: Creating the Latenode Scenario with AI based customer support

Now, we'll set up the automation scenario in Latenode with Ai for customer service, incorporating chatbot customer support.

Authorization Token:

• Go to Authorizations and create a new authorization for Google Workspace, specifically for Gmail.

• Use the email address where your users will send their questions to obtain an access token.

Create Scenario:

• Navigate to scenarios and create a new one.
• Add the first node, "Create Thread."

Creating a Thread:

• Inside the "Create Thread" node, insert your OpenAI API key in the authorization column and click "run once" to generate a thread ID.

Schedule Trigger:

• Add a schedule trigger node and set the activation period and timezone.

Get List of Messages:

• Add an HTTP-request node to fetch a list of unread messages from your mailbox.
• Configure the node with the appropriate URL, replacing {userId} with your email address, and add "?q=is:unread" to retrieve only unread messages.
• Set the method to "GET" and include the authorization token.

Iterator:

• Add an iterator node to process each message in the array obtained from the previous node.

Get Messages:

• Add another HTTP-request node to retrieve the data from each unread message.

‍

• Configure the node with the appropriate URL using the message ID obtained from the iterator.

JavaScript Node:

• Add a JavaScript node to process sender emails and email bodies.
• Employ the provided code or customize it according to your requirements, enhancing AI-powered customer support.
  
  Here’s the code that you can put here:

export default async function run({ execution_id, input, data }) {
  let headers = JSON.parse(data["{{6.body.payload.headers}}"]); // Corrected the data path
  // from
  const email_from = headers.find((item) => item.name === "From");
  const email_from_value = email_from.value.match(/<([^>]+)>/)[1]; // Use regex to extract email

  // snippet (content)
  const email_content = data["{{6.body.snippet}}"]; // Corrected the data path

  return {
    "email_from": email_from_value,
    "email_content": email_content
  };
}

• Ensure to replace the number of the previous node accordingly.

Create Message:

• Add a ChatGPT node to generate a response message using the extracted email body.

• Insert your OpenAI API key and Thread ID.

Create Run:

• Add a node to send the message to the assistant.
• Insert OpenAI API key, Thread ID, and Assistant ID.
• Choose the model version and set "true" in the Retrieval field.

Gmail Node:

• Select the Gmail "send message" node and configure it with the necessary details.
• Use the email_from variable from the JavaScript node in the "To" field.
• Include the email body from the Create Run node.

Modify Messages:

• Add a request to mark messages as read.

Choose raw format of the body and insert this:

{

  "removeLabelIds": ["UNREAD"]

}

• Configure the node with the appropriate URL and method.
• Include the necessary key-value pairs.

Save the scenario, deploy it, and here you are! You have an AI powered customer service assistant which can answer customer emails for you!

Conclusion

‍

In conclusion, leveraging AI in customer service and chatbot customer support can revolutionize customer interaction. Integrating AI for customer support streamlines processes, while AI-powered customer service enhances satisfaction. Adopting AI customer care enables accurate issue resolution. AI tech support offers round-the-clock assistance. Deploying an AI assistant to handle customer emails amplifies efficiency. 

For assistance, reach out to Latenode's Discord community. 

Now, deploy your scenario and enjoy the AI support benefits of automated email response!

‍