About
Greg Sollenberger, FCAS, CSPA
I am a property and casualty actuary and data scientist with experience in insurance pricing, predictive modeling, ratemaking, and model governance.
Much of my technical work has focused on generalized linear models and other predictive methods, translating model results into practical rating structures, and building workflows that make actuarial analysis easier to explain, validate, reproduce, and implement.
I work primarily in R and have also worked with a broader modeling and production stack including gradient boosting, regularized regression, APIs, and related data-science tooling.
Sollenberger Actuarial Technologies
Sollenberger Actuarial Technologies is a vehicle for publishing open-source actuarial software, demonstrations, and related technical work.
The projects are motivated by problems that arise repeatedly in actuarial practice: rating systems that are difficult to audit, rate revisions that lose their history in spreadsheets, models that are hard to translate into implementation, and analytical workflows that are difficult for someone else to reproduce.
The aim is not to replace actuarial judgment with software. It is to make the technical machinery around that judgment clearer and easier to work with.
Open-source work
Current projects include:
ratingtables, a table-driven insurance rating engine;raterevision, tools for reproducible rate-table revision workflows; andaglmtools, a development-stage project for interpretable regularized GLMs.
I am particularly interested in feedback from people trying to use these tools on real rating structures and implementation workflows. Practical edge cases are often the most useful source of ideas for improving the software.
Elsewhere
The projects and views presented here are my own and are not presented on behalf of any employer.