Hesper Nightall-Kariuki

Prepare quarterly loss-development triangles for personal auto and homeowners lines covering $410M in earned premium. Run chain-ladder and Bornhuetter-Ferguson reserve estimates; reconcile paid and incurred indications and flag outliers for the reserving actuary. Built a Python pipeline (pandas, SQL Server) that refreshes triangles from the claims warehouse, cutting quarter-close prep from 5 days to 2. Support Connecticut and Massachusetts auto rate filings: trend selections, on-leveling premium, and exhibit preparation for regulators. Analyze territory relativities with GLMs in R; findings fed a 2025 rate revision that improved the homeowners loss ratio 3.1 points. Maintain the catastrophe load exhibit and document assumptions for internal audit and the appointed actuary. Present monthly large-loss and frequency/severity dashboards (Power BI) to underwriting managers.

Validated mortality and lapse assumptions for a term-life block of 62,000 policies using experience studies in Excel and SQL. Automated a VBA macro for monthly in-force summaries, saving the valuation team about 6 hours per month. Presented an intern capstone on lapse sensitivity under rising interest rates to the chief actuary.

Tutored 15+ students per week in calculus, probability, and statistics; prepared exam review sessions for 100-level courses.

GPA 3.7/4.0; Dean's List (6 semesters). Coursework: Probability, Financial Mathematics, Loss Models, Regression & Time Series, Life Contingencies, Database Systems.

Passed 2023

Passed 2023

Sitting Spring 2027

Complete

Organized 4 employer panels and an exam-prep study group of 40 members.

Helped prepare 60+ returns for low-income Hartford-area households.

Fit Poisson and negative-binomial models on 1.2M open-source policy records; documented model diagnostics on GitHub.

Built an R Shiny app that bootstraps chain-ladder reserves to show uncertainty ranges for non-actuarial audiences.