The Story Behind Mostly Optimal
After 11 years as an actuary working across life insurance, analytics, commercial casualty, and brokerage, I kept seeing the same gap: the analytical tools available to insurance buyers are decades behind what carriers use internally. So I built the tool I wished existed.
My career has spanned the insurance industry: from building dynamic capital models at a Fortune 100 carrier, to modernizing analytical platforms at Verisk, to optimizing multi-million-dollar insurance programs as a commercial broker. Each role reinforced the same observation: what happens to a single company over time looks nothing like the industry average.
When I discovered the mathematical framework behind this observation, everything clicked. A quick Excel prototype turned into a Python simulation engine, then a research paper, then the open-source platform you see here. Hundreds of hours later, Mostly Optimal is ready for the community to use, challenge, and build on.
Vision
The insurance industry treats risk as a cost to minimize. It should be a lever for growth. This framework is my attempt to prove that and give to every insurance buyer the analytical tools that carriers have kept to themselves.
Values
I built this project on three convictions:
- Open-mindedness: The best ideas often come from questioning what everyone takes for granted.
- Analytic rigor: Intuition matters, but it deserves to be tested.
- Better decisions: The point of all this analysis isn't the analysis itself. It's making decisions you can stand behind.
Yours in chaos,
Alex Filiakov, ACAS