Analytics & Insurance Operations
13+ years building analytics infrastructure, leading high-performing teams, and turning complex insurance data into decisions that move the business. Based in Addison, IL.
Scroll to explore
About
I've spent my career at the intersection of insurance domain expertise and technical analytics — managing regulatory compliance at national scale, building reporting infrastructure that VP-level leaders rely on, and modernizing how analytics teams work. My background spans personal lines pricing, state product management, regulatory reporting, and business performance analysis across books of business exceeding $3B in gross written premium.
I believe the best analytics work is invisible to the people who use it — the report is just there, it's right, and it answers the question before anyone has to ask twice.
Core Skills
SQL, R, Python, SAS — workflow automation, regulatory reporting, pipeline construction, funnel analysis
SQL Server (SSMS / DBeaver), Power BI / DAX, SSRS, VS Code, R Studio, Domino, SAS, Excel, GitHub
Claude API, Plaid API, Ollama (local LLM), LLM-assisted data workflows, prompt engineering, API integration
Personal lines insurance, pricing strategy, risk management, state product management, regulatory compliance
Team building, analyst mentoring, 1:1 coaching, cross-functional stakeholder management, executive reporting
Executive dashboards, VP-ready package design, business unit reviews, KPI frameworks, field-to-senior delivery
Personal Project
Full-stack personal finance tracker with local AI analysis — built end to end
I wanted a finance tracking tool that gave me real analytical insight into my spending — not just categorized totals, but actual pattern recognition and trend analysis. Every commercial option either required trusting a third party with sensitive financial data or delivered surface-level insights that didn't answer the questions I actually had. I decided to build my own.
The application integrates the Plaid API to pull transaction data from connected financial accounts into a local data pipeline. That data is cleaned, categorized, and structured through a custom ingestion layer before being passed to Ollama — a locally hosted large language model — for on-device AI analysis. Nothing leaves the machine. The Claude API supported the development workflow itself, helping iterate on architecture decisions and code structure throughout the build.
Building this end to end reinforced how much of real-world API work is about handling edge cases in data ingestion — Plaid's transaction schema is inconsistent across institutions, and building a reliable normalization layer took meaningful iteration. Running LLM inference locally via Ollama was a deliberate privacy choice that also forced me to think carefully about prompt structure and context management in ways that cloud API calls abstract away. The project made me significantly more comfortable working across the full stack from data ingestion through AI-assisted output.
Experience
Contact