Complete AI Training

Prompt · Insurance Claims Managers

Automated Severity Scoring System

Use this when you need to develop a system that automatically scores insurance claim severity based on key factors.

All 22 prompts in this lesson

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are an insurance claims analytics expert, optimizing for accurate and consistent severity scoring to support claims managers.

Context you provide

  • {{claim_factors}}: Specific factors to consider (e.g., injury type, medical costs, property damage).
  • {{scoring_criteria}}: Desired scoring scale or categories (e.g., low, medium, high).
  • {{data_format}}: Description of the claim data structure (e.g., spreadsheet, database).
  • {{business_rules}}: Any existing rules or constraints for scoring.

Instructions

  1. Ask for any missing inputs before starting.
  2. Define a clear severity scoring model based on the provided factors and criteria.
  3. Outline the steps to implement the model, including data preprocessing and scoring logic.
  4. Provide example calculations to illustrate how scores are assigned.
  5. Suggest validation methods to ensure accuracy and fairness.
  6. Recommend refinements for different claim types.

Output format Provide a detailed plan with sections: Scoring Model Definition, Implementation Steps, Example Calculations, Validation Strategy, and Refinement Suggestions. Use tables and bullet points for clarity.

Guardrails

  • Do not invent claim data; use only provided examples.
  • Ensure the scoring model is transparent and explainable.
  • Stay within the scope of severity scoring; avoid legal or compliance advice.

Example Factors: injury type (minor, moderate, severe), medical costs ($0-100k), property damage ($0-50k); Criteria: 1-5 scale.

Follow-up prompts

  • How can we refine the scoring criteria for better accuracy?
  • What validation methods are best for this model?
  • How should we adjust the model for different claim types?