Home

I'm Joe, an Actuarial Data Scientist focused on Commercial P&C pricing. I build, implement, and integrate competitive pricing models, monitoring tools, dashboards and deployment patterns that have a proven track record of pricing and profitability for the lines of business that I serve. My models, rooted in my background both in actuarial science and data science, support real-world decision making across key stakeholders that produce actionable, repeatable, and dynamic improvements applied to niche and key markets.

What distinguishes me from others in my field is that my modeling philosophy was borne through experience that combines a diverse background in both actuarial science and data science but also from the lines of business and markets that I have served. I have worked with personal lines and commercial lines in super regional carriers and a global insurance carrier.

My multidisciplinary background has allowed me to leverage my experience to productively work and provide actionable data-driven advice with multiple competing internal business needs, objectives, and priorities. I have worked in environments where pricing models had to survive more than a validation metric: messy insurance data, actuarial review, product strategy, underwriter judgment, market pressure, regulatory scrutiny, production deployment, and post-launch monitoring. That experience forms the foundation of my data science practice: the best pricing models are not just technically strong- they are trusted, rational, monitored, dynamic, and responsive to contemporary market conditions.

At the heart of what I do is the fundamental belief that data work is not beneath the modeler; it is the model. Credibility beats cleverness. A pricing model is a business system, not just a math artifact. User feedback is model monitoring. And insurance pricing should be managed continuously, because no model is ever finished.

This website showcases a sample of my work product that displays not just my talent- but the depth and breadth of my experience that will add value to your insurance organization. If this intrigues you, let's connect on Linkedin and see if I can't help your organization leverage its data better to produce better pricing and profitability.

Featured Projects

Note: Project links are being added; summaries below describe each project at a high level.

Unless noted otherwise, each project includes underlying code, documentation, and presentation materials tailored for both technical and non-technical audiences.

  • Presentations & Teaching



  • Data Analysis

    • Close Ratio Shiny Dashboard (Python, Shiny) Interactive Shiny app to explore quote-to-bind close ratios by risk attributes. Used by underwriting, distribution and product to identify underperforming segments and refine quoting strategy.
    • Pricing Model Interpretability Dashboard (Python, Shiny) Dashboard for explaining GLM and ML pricing models to business users such as product and underwriting.


  • Computer Vision Models

    • Vehicle Damage Classifier (Python, FastAI) Proof of concept, simple computer vision model that can classify a vehicle as damaged or not from a photo. Learning project to understand web scraping, FastAI and computer vision. Could be expanded to predict car repair costs with proper training data.


  • Classification Models

    • Close Ratio Analysis (Python) Binary classifier predicting likelihood of bind at quote time using producer and risk attributes.
    • Claims Fraud Detection (Python) Fraud detection model to prioritize suspicious claims for SIU review.
    • Industry Code Labeling (AI, LLM, Sentence Embeddings, Neural Net, Python, Flask, Shiny) Hierarchical machine learning model for predicting 2022 NAICS industry codes from company descriptions.


  • Regression Models

    • Insurance Pricing Models: GLM Methods (Python) Traditional GLM-based pricing framework with exposure, frequency, and severity components.
    • Insurance Pricing Models: Machine Learning Methods (Python) Machine learning models benchmarked against GLMs for improved rate segmentation.


  • Model Delivery & Deployment

    • Insurance Pricing Model API (Python, Flask, Shiny) UI implemented via Shiny for harvesting user inputs, Flask API for predicting price from provided inputs and Shiny dashboard to display pricing model results along with interpretability exhibits.
    • Industry Code Labeling (Flask, Shiny, AI, LLM, Sentence Embeddings, Neural Net, Python) Hierarchical machine learning model for predicting 2022 NAICS industry codes from company descriptions.


  • Post-Production Monitoring

    • Insurance Pricing Model (Python) Monitoring suite tracking lift, drift, calibration, and stability of deployed pricing models. Intended to be used by business partners like underwriting and product to understand model performance after launch.

Contact

Connect with me on LinkedIn

Full resume and additional project details available on LinkedIn or upon request.

Skills

Technical

  • Languages: Python, R, SQL
  • Pricing & Predictive Modeling: Commercial P&C pricing models, GLMs, Tweedie modeling, frequency/severity/pure premium/loss ratio models, gradient boosting, GAMs, deep learning, regression, classification, cross-validation, regularization, model validation, and monitoring
  • Tools & Platforms: Jupyter, VS Code, Git, Hugging Face, AWS SageMaker, AWS Athena, Flask, Shiny, JIRA, Excel, VBA, Power BI
  • Data & Production Analytics: SQL Server, large-scale policy and claims data, pricing data pipelines, feature engineering, data reconciliation, deployment support, dashboards, and model monitoring

Domain & Leadership

  • Commercial P&C Pricing: Commercial Auto, General Liability, Commercial Property, Businessowners
  • Pricing Modernization: Rate plan refinement, segmentation strategy, model-informed pricing, portfolio analytics, underwriting analytics, and pricing execution
  • Model Lifecycle Management: End-to-end ownership across data foundation, model development, actuarial review, regulatory support, deployment, monitoring, governance, and continuous improvement
  • Leadership: Player-coach modeling leadership, mentoring analysts/modelers, guiding technical quality, partnering with actuarial/product/underwriting/technology teams, and connecting model output to pricing and underwriting execution

Opportunities I'm interested in

I'm interested in hands-on leadership roles in Commercial P&C pricing, portfolio management, and underwriting analytics- roles where I can help set modeling strategy, guide technical execution, mentor modelers and analysts, and stay close enough to the work to ensure models succeed in production.

I am especially drawn to player-coach opportunities where I can help build, scale, or modernize a pricing and analytics function: strengthening data foundations, developing credible and production-ready pricing models, improving model governance and monitoring, partnering across actuarial, product, underwriting, and technology teams, and connecting model output to underwriting execution and portfolio performance.

Location Preferences: Columbus, OH; open to remote roles.

If you have an opportunity you'd like to share with me, please reach out.

Education & Credentials

Education

  • Otterbein University — B.S. in Actuarial Science, May 2013
  • Texas A&M University — M.S. in Statistical Data Science, December 2026

Actuarial Exam Progress

Progress toward CAS credentials. Completed items are marked below.

ACAS

  • ✅ Exam 1: November 2011
  • ✅ Exam 2: July 2012
  • ✅ VEEs: May 2013
  • ⬜ MAS-I: Sitting October 2026