n8n Workflow: Score Telematics Driving Risk with Claude and Adjust Insurance Premiums

This n8n workflow automates the process of analyzing telematics data to assess driving risk using an AI agent (Claude) and then adjusts insurance premiums based on the risk score. It also handles communication by sending emails for low-risk drivers and Slack notifications for high-risk cases.

What it does

This workflow performs the following key steps:

1. Triggers on Schedule: The workflow is initiated on a predefined schedule (e.g., daily, weekly).
2. Fetches Telematics Data: It makes an HTTP request to an external API to retrieve telematics driving data.
3. Analyzes Driving Risk with AI: An AI Agent, powered by the Anthropic Claude Chat Model, processes the telematics data to score the driving risk. A Structured Output Parser ensures the AI's response is in a usable format.
4. Processes AI Output: The AI's risk assessment is then prepared for conditional logic.
5. Evaluates Risk Score: An "If" node checks the calculated risk score.
   • Low Risk: If the risk is low, it sends an email via Gmail to the driver with a positive update.
   • High Risk: If the risk is high, it sends a Slack notification to an internal team for review.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• Anthropic API Key: For the Anthropic Chat Model (Claude).
• External Telematics API: An API endpoint that provides driving data.
• Gmail Account: Configured as a credential in n8n for sending emails.
• Slack Account: Configured as a credential in n8n for sending notifications.

Setup/Usage

1. Import the Workflow:
   • Download the workflow JSON.
   • In your n8n instance, click "Workflows" in the left sidebar.
   • Click "New" -> "Import from JSON" and paste the workflow JSON.
2. Configure Credentials:
   • Anthropic Chat Model: Add your Anthropic API key as a credential for the "Anthropic Chat Model" node.
   • HTTP Request: Configure the "HTTP Request" node with the URL and any necessary authentication for your telematics data API.
   • Gmail: Set up your Gmail credentials for the "Gmail" node.
   • Slack: Set up your Slack credentials for the "Slack" node.
3. Customize Nodes:
   • Schedule Trigger: Adjust the schedule to your desired frequency (e.g., daily, weekly).
   • HTTP Request: Modify the URL, headers, and body as needed to fetch your specific telematics data.
   • AI Agent: Review the prompt for the AI Agent to ensure it correctly interprets your telematics data and outputs the desired risk score format.
   • Structured Output Parser: Ensure the schema matches the expected output from the AI Agent.
   • Edit Fields: Adjust the fields being set if your data structure differs.
   • If Node: Modify the condition to define what constitutes "low risk" and "high risk" based on the AI's output.
   • Gmail: Customize the recipient, subject, and body of the email for low-risk drivers.
   • Slack: Customize the channel, text, and any other details for the high-risk notification.
4. Activate the Workflow: Once configured, activate the workflow to start processing telematics data on schedule.