Torleif - Actuary AI Skill

Turns a messy claims history into a defensible reserve number, rate indication, or board-ready explanation of what the company doesn't yet know for certain.

Torleif is a Senior Actuary persona that builds loss development triangles and checks them for data quality before computing a single factor, chooses chain-ladder or Bornhuetter-Ferguson deliberately based on triangle maturity, separates indicated from carried reserve and diagnoses the driver behind any gap, grounds rate indications in exposure and trend rather than last year's rate plus inflation, selects mortality and morbidity tables to match the actual population, weights experience with a stated credibility formula, and translates every result into a best estimate plus range for a board with no actuarial background assumed.

Who it's for

Insurance companies, reinsurers, and benefits programs that need reserve, pricing, or credibility analysis with the mechanics shown, not a black-box number, and a board-ready plain-language explanation of what the range means.

Key capabilities

• Builds loss development triangles and states which diagonal or factor looks unstable and why
• Chooses chain-ladder or Bornhuetter-Ferguson deliberately based on triangle maturity and stability
• Builds rate indications from earned exposure, developed/trended losses, and expense and profit load
• Selects mortality/morbidity tables matched to the actual population and blends in credible company experience
• Weights pool experience with a stated credibility formula and translates every result into a best estimate plus range

How to use it

Paste Torleif's SKILL.md into your Claude Project Instructions (or any AI system prompt), then describe your reserve, pricing, mortality table, or credibility problem. Works with Claude, ChatGPT, and any AI chat. Under 2 minutes to install.