Insurance Data Analyst
Most analysts study insurance from the outside. I understand how claims are lodged, reserves are set, and regulatory frameworks shape every data point — and I build the pipelines, models, and RAG tools to work with them.
End-to-end insurance data warehouse built on 1.7M real flood insurance claims and 1M policies from FEMA's NFIP, ingested via REST API. Medallion Architecture pipeline producing loss ratios, claims triangles, and portfolio-level KPIs across five US states.
A six-model machine learning benchmark on 10,000 bank customers — investigating not just which algorithm wins, but how to meaningfully improve recall on a real-world imbalanced dataset using SMOTE, sample weighting, and threshold tuning.
Tweedie regression pipeline estimating Pure Premium across 677k+ motor insurance policies — with reactive Streamlit dashboard, model evaluation, and feature importance analysis.
A RAG-powered decision-support assistant for Australian Workers Compensation queries, built by a practising Claims Advisor at Suncorp Group. Grounds LLM answers in authoritative legislation and policy documents with inline source citations and a 20-question evaluation harness.
When inflation rises, insurance claims get more expensive — but the bill doesn't arrive until months later. Using ABS, RBA, and APRA data, I built an analytics system to detect economic warning signals before they hit the books: lead-lag correlations, predictive models, and a live stress-test dashboard for four macro scenarios.
Open to opportunities — Insurance Data Analyst & BI Analyst roles in insurance and financial services