Sollenberger Actuarial Technologies
Open-source actuarial software for transparent, reproducible insurance work
Sollenberger Actuarial Technologies is a vehicle for publishing open-source software and technical work focused on insurance rating, pricing, predictive modeling, and implementation.
The goal is practical: make actuarial work easier to understand, test, reproduce, review, and carry into production.
Actuarial technology, built in the open
The projects published here grow out of recurring actuarial problems: representing complex rating plans clearly, preserving an auditable chain from current to proposed rates, fitting useful predictive models, and making analytical work easier to validate and implement.
The emphasis is on tools that are:
Transparent
Calculations and model structures should be inspectable rather than hidden behind opaque interfaces.
Reproducible
Important analytical and rating decisions should be expressible in code, version controlled, and repeatable.
Practical
Software should fit real actuarial workflows, including Excel interaction, implementation review, auditing, and production use.
Current projects
ratingtables
A lightweight, table-driven R rating engine for executing insurance rating plans from normalized factor tables and an ordered calculation specification.
It supports coverage-specific specifications, entity-level rating and aggregation, rate-set selection, validation, and optional step-by-step calculation traces.
Learn more · GitHub · CRAN
raterevision
An R package for managing rate-table revisions around a table-driven rating workflow.
It supports exporting and importing rate tables through Excel as well as programmatic rate changes that preserve a clear record of how proposed rates were produced.
Learn more · GitHub · Demos
aglmtools
A development-stage R project for interpretable regularized generalized linear modeling, including cumulative ordered encodings and actuarial-oriented handling of numeric and categorical predictors.
Technical focus
The software and technical work here are informed by experience in areas including:
- property and casualty insurance pricing and ratemaking;
- predictive modeling and generalized linear models;
- rating-plan design, implementation, and validation;
- model documentation, governance, and reproducibility;
- retention and profitability modeling;
- R-based actuarial workflows and tooling.
Interested in using the software?
Feedback from actuaries, analysts, implementation teams, and other users is especially useful. I am happy to discuss the packages, implementation questions, difficult rating structures, modeling problems, or ideas for extending the tools.