Understanding the AI Insurance Model

At the core of AI Insurance’s platform is a flexible, object-based data model designed to reflect how insurance policies are structured in the real world. This page explains how rating data is structured within the AI Insurance platform, giving you a clear understanding of how we organize and store the information that drives underwriting, pricing, and claims analysis.

What Is Rating Data?

Rating data is the information used to calculate the price—or premium—of an insurance policy. It helps underwriters and rating engines evaluate how risky something is to insure, and how much it should cost. This includes:

• What kind of policy it is
• What is being insured (e.g., buildings, vehicles, people)
• Specific details about those things—like size, structure, revenue, or safety features

---

How AI Insurance Organizes This Data

Think of a policy like a folder that holds everything about an insurance contract. Inside that folder, rating data is organized into two main levels:

1. Policy-Level Rating Data

This includes information that applies to the entire policy—regardless of how many items (or “risks”) it covers. This is like the cover sheet of the folder—it applies to everything inside.

Examples:

• Policy Type (e.g., “Claims Made”)
• Brokerage (e.g., NFP)
• Program name, Effective/Expiration Dates, etc.

2. Risk-Level Rating Data

Each item being insured—called a risk—has its own set of data fields.

For example, if you’re insuring three buildings, each one is treated as a separate risk object, each with its own details:

• Number of stories
• Construction material (e.g., stone, wood)
• Fire suppression system (Yes/No)
• Estimated replacement cost
• Annual revenue (for business risks)

This level of detail allows the platform to rate each risk independently based on its features.

Example: Insuring 3 Buildings

Let’s say the policy is for Resident’s Group LLC, which owns 3 buildings. Here’s how that data would be structured in AI Insurance:

Policy: ABC-1234

• Policy Type: Claims Made
• Brokerage: NFP
• Insured Entity: Resident’s Group LLC
• FEIN: [XX-XXXXXXX]
• Annual Revenue: $X,XXX,XXX

Covered Risks:

• Building 1

  • 3 stories
  • Stone masonry
  • Fire suppression system: Yes

• Building 2

  • 4 stories
  • Joisted wood
  • Fire suppression system: Yes

• Building 3

  • 6 stories
  • Stone masonry
  • Fire suppression system: No

In AI Insurance, each building has its own digital “card” inside the policy folder. This enables precise pricing, reporting, and analytics.

How the Data Model Works for Quotes and Policies

The same object-based structure that organizes rating data also powers the entire insurance lifecycle—from initial quotes to active policies. Here’s how it works:

Quotes: The Foundation

When creating a quote, the data model captures the same two-level structure:

1. Quote-Level Data:

• Quote type and effective dates
• Brokerage and program information
• Overall terms and conditions

2. Risk-Level Data:

• Individual risk characteristics for pricing
• Coverage limits and deductibles per risk
• Rating factors and modifiers

This structure allows rating engines to calculate premiums for each risk independently, then aggregate them into a total quote premium.

From Quote to Policy

When a quote is bound and becomes a policy, the data model maintains the same structure but adds policy-specific information:

Policy-Level Additions:

• Policy number and binding information
• Premium payment schedules
• Policy endorsements and modifications

Risk-Level Continuity:

• The same risk objects carry forward from quote to policy
• Individual risk premiums and coverages are preserved
• Risk-specific endorsements can be added

Example: Quote to Policy Flow

Step 1: Quote Creation

• Quote Q-2024-001 for Resident’s Group LLC
• 3 buildings with individual risk assessments
• Each building rated independently based on its characteristics

Step 2: Quote Binding

• Quote becomes Policy ABC-1234
• All risk-level data preserved
• Policy-level terms finalized

Step 3: Policy Management

• Individual buildings can be modified, added, or removed
• Risk-specific endorsements applied
• Claims linked to specific risk objects

This seamless flow from quote to policy ensures data consistency and enables precise tracking throughout the insurance lifecycle.

Why This Data Model Matters

The AI Insurance data model isn’t just about organizing information—it’s about enabling better insurance decisions through structured, granular data. Here’s what this approach delivers:

Precision in Pricing

By treating each risk as an independent object with its own characteristics, underwriters can apply the most accurate rating factors to each individual item. A 6-story building without fire suppression gets rated differently than a 3-story building with full sprinkler systems, even if they’re on the same policy.

Flexibility in Policy Structure

The folder-and-cards approach mirrors real-world insurance relationships. Whether you’re insuring one building or fifty, the same data structure applies. Add or remove risks without disrupting the overall policy framework.

Enhanced Analytics and Reporting

With each risk object containing its own detailed data, you can generate insights at multiple levels:

• Policy-level reports show overall portfolio performance
• Risk-level analysis reveals which specific items drive losses or profitability
• Comparative analytics help identify patterns across different risk types

Streamlined Workflows

This object-based approach supports modern insurance workflows:

• Rating engines can access individual risk data for automated pricing
• Claims systems can link incidents to specific risk objects
• Audit processes can verify coverage against individual risk characteristics

AI Insurance’s data model organizes rating data into policy-level and risk-level components, the platform provides the foundation for accurate pricing, comprehensive reporting, and data-driven decision making—all while maintaining the intuitive structure that insurance professionals expect.