Loading repository data…
IAA-AITF / repository
Actuarial-AI-Case-Studies
A curated collection of actuarial AI case studies covering machine learning, generative AI, agentic AI, and more, designed to support actuaries in applying AI to real-world problems.
#actuarial-science#artificial-intelligence#case-studies#generative-ai#machine-learning
Live repository statistics
★ 41Stars
⑂ 11Forks
◯ 1Open issues
◉ 41Watchers
README preview
Actuarial AI Case Studies
A curated collection of actuarial AI case studies, created by the International Actuarial Association's (IAA) AI Task Force. This repository demonstrates how AI — encompassing machine learning, generative AI, agentic AI, and more — can be applied to real-world actuarial problems.
Featured Case Studies
| Case Study | Topics | Level |
|---|---|---|
| Car Damage Classification and Localization | Fine-Tuning, Vision Models, Structured Outputs | Advanced |
| Data Analysis Multi-Agent System | Multi-Agent Systems, Agent Orchestration, Automated Reporting | Advanced |
| GenAI-Driven Market Comparison | RAG Pipelines, Structured Outputs, Document Analysis | Advanced |
| Claim Cost Prediction with LLM-Extracted Features | LLMs, Feature Engineering, Claims Severity | Advanced |
Browse the full catalog of case studies (including external references) in the Case Studies Directory.
Repository Structure
Actuarial-AI-Case-Studies/
├── case-studies/ # Case studies organized by year
│ ├── README.md # Catalog index of all case studies (hosted + external)
│ ├── 2025/ # Hosted notebooks, data, and documentation
│ └── 2026/ # Hosted notebooks, data, and documentation
├── templates/ # Templates for new case study submissions
├── CONTRIBUTING.md # Contribution guidelines
└── LICENSE # MIT (code) + CC BY 4.0 (content)
Getting Started
Each case study is self-contained in its own directory with a README.md, a Jupyter notebook (.ipynb), and a requirements.txt. To run a case study locally:
git clone https://github.com/IAA-AITF/Actuarial-AI-Case-Studies.git
cd Actuarial-AI-Case-Studies/case-studies/<year>/<case-study-name>
pip install -r requirements.txt
jupyter notebook <notebook-name>.ipynb
Alternatively, open the .ipynb files directly in Google Colab or Kaggle.
Contributing
We welcome contributions from actuaries, data scientists, and AI practitioners. Please review the CONTRIBUTING.md for submission guidelines, or use one of the provided templates to get started.
License
- Content (case studies, articles, documentation): Creative Commons Attribution 4.0 (CC BY 4.0)
- Code (scripts, Jupyter notebooks): MIT License
- Third-party materials: Distributed under their respective licenses, as indicated.
Contact
For questions or suggestions, open an issue or contact us via email.