n8n Workflow: Analyze Sales Calls with GPT-4 and Deliver Coaching Feedback via Slack & Google Sheets

This n8n workflow automates the process of analyzing sales call transcripts using GPT-4, generating coaching feedback, and delivering it to a sales representative via Slack, while also logging the feedback in a Google Sheet.

What it does

This workflow is designed to streamline the post-call analysis and coaching process for sales teams.

1. Receives Sales Call Data: The workflow is triggered by an incoming webhook, expecting a sales call transcript and the sales representative's name.
2. Analyzes Transcript with GPT-4: It sends the sales call transcript to a GPT-4 model (via an HTTP Request node) to generate detailed coaching feedback.
3. Formats Feedback: The generated feedback is then processed and formatted into a structured message.
4. Posts to Slack: The coaching feedback is sent as a direct message to the specified sales representative on Slack.
5. Logs to Google Sheets: The entire interaction, including the transcript and the generated feedback, is logged in a Google Sheet for record-keeping and further analysis.
6. Responds to Webhook: Confirms successful processing back to the triggering system.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• OpenAI Account & API Key: For accessing GPT-4 (or a compatible LLM API). The HTTP Request node will need to be configured with the appropriate endpoint and authentication for your chosen LLM.
• Slack Account & Bot Token: To send messages to sales representatives. You'll need to set up a Slack credential in n8n.
• Google Sheets Account: To log the call data and feedback. You'll need to set up a Google Sheets credential in n8n.
• Google Sheet: A Google Sheet with appropriate columns to store the sales call data (e.g., "Sales Rep Name", "Transcript", "Coaching Feedback", "Timestamp").

Setup/Usage

1. Import the Workflow:

   • Download the provided JSON file.
   • In your n8n instance, click "Workflows" in the left sidebar.
   • Click "New" and then "Import from JSON".
   • Paste the JSON content or upload the file.

2. Configure Credentials:

   • HTTP Request (for LLM): Open the "HTTP Request" node. Configure it to connect to your LLM API (e.g., OpenAI's GPT-4 endpoint). You'll need to specify the API URL, request method (likely POST), headers (e.g., Authorization: Bearer YOUR_OPENAI_API_KEY), and the request body containing the prompt and the sales call transcript.
   • Slack: Open the "Slack" node. Select or create a Slack credential. This typically involves providing a Bot User OAuth Token.
   • Google Sheets: Open the "Google Sheets" node. Select or create a Google Sheets credential. This usually involves OAuth2 authentication to grant n8n access to your Google Sheets.

3. Configure Nodes:

   • Webhook: The "Webhook" node is the trigger. Copy its URL once the workflow is active. This URL is where your external system (e.g., a call transcription service) should send the sales call data.
   • Edit Fields (Set): This node is currently empty. You might use it to rename or transform incoming data fields before sending them to the LLM or other services.
   • Code: This node is currently empty. It's a placeholder for custom JavaScript logic, which could be used to refine prompts, parse LLM responses, or format data more complexly.
   • Google Sheets: Configure the "Google Sheets" node:
     • Specify the "Spreadsheet ID" and "Sheet Name" where the data should be logged.
     • Map the incoming data fields (Sales Rep Name, Transcript, Coaching Feedback) to the corresponding columns in your Google Sheet.
   • Slack: Configure the "Slack" node:
     • Set the "Channel" or "User ID" to send the message to the sales representative. You can use an expression to dynamically get the user ID from the incoming webhook data.
     • Define the "Text" of the message, incorporating the coaching feedback generated by the LLM.
   • Respond to Webhook: This node will send a confirmation back to the system that triggered the workflow. You can customize the response body if needed.

4. Activate the Workflow: Toggle the workflow to "Active" in the top right corner of the n8n editor.

5. Test the Workflow: Send a test request to the Webhook URL with sample sales call data (e.g., {"salesRepName": "John Doe", "transcript": "Customer said X, I responded Y..."}) to ensure everything is working as expected.