Developing Automation & GenAI functionality from scratch (building the team that has architected & executed on real problems - Submission Intake, Automated data extraction for Quote, Claims Intake, Claims triage, automated data comparison). Brought in a modern suite of vendor solutions that work on Fixed Form, Semi Structured, & Unstructured data, partnering with the Human in the Loop developers to create Arch's first HITL UI.

Oversee all of Hippo data science (insurance, marketing, growth, and internal ops)• Forged good working relationships with Business Units, especially those skeptical of Data Scienceo Developed strong Buy-In, Identified Key players to help increase visibility of our work• Built out lean teams across 3 different offices & 2 countries to meet the analytics needs of our BU partners o Developed & Mentored Mid-Career Employees to increase productivity, elevate high performing individuals, and retain top talent. o Refined & streamlined hiring processo Cultivated a solutions-oriented team that focuses on adding real business value to every product they touch (reframed team from Research to Development)• Oversaw & supported employees to achieve rapid results, high quality prototypes, and many dashboards. These efforts directly produced $8M (~3% of Combined Ratio) savings in my first year• Collaborated with ML Ops to create both ML production architecture & Data Science research platforms (AWS)o Personally built out high performance data science infrastructure for use by data scientists• Reviewed & assessed 3rd party Data Science SAAS & 3rd Party Data Vendors for use by Hippo. Guided development of a set of tools to determine marginal value of 3rd party offerings. • Supported a large Actuarial-Data Science collaboration to rebuild our Rate Plan using GLMs by committing DS resources and serving as the technical consultant• Worked with many BU leaders to understand their needs (vs wants), negotiate acceptable solutions & KPIs, prioritized our efforts to ensure progress, and ultimately, add value.

Leads a growing team of 17 data scientists & data engineers working across countries (USA & Poland) and built this team from the ground-up. We're primarily an internal consulting group responsible for advanced data science & engineering and modernizing our technical platform. Highlights include NLP and visualizations of the Voice of our Customers, Advanced Entity Resolution, Unbiasing of data (protected classes), Machine Learning for Rating, and Fast Prototyping of numerous Commercial and Personal Lines projects.

Lead a team of 17 data scientists & data engineers working across countries (USA & Poland) and built this team from the ground-up. We're primarily responsible for advanced data engineering and modernizing our technical platform. Highlights: Using a combination of complex matching algorithms, natural language processing, and optimization, we deanonymized address information to create new modeling data sets. Overseen the transition from SAS to Spark/GPUs on AWS and a wide roll out of our environment to 40 users (soon to be 300). Continued progress towards building/releasing Regulator Approved Machine Learning based Pricing.

Lead efforts to create new products surrounding Micro-Territories, Commercial Auto Symbols, Liability Frequency Trend projection as well as providing actionable recommendations for updating key assets by using ML algorithms to identify which data assets will produce the most revenue. My role has involved management of, and participation on, a small team (3 – 5). I have been heavily involved in every phase from data preparation (cleaning, merging, feature engineering, etc) to creating the final predictive models. Most recently, I have conceived of and spearheaded the creation of a Property Vacancy Detection product with a projected revenue of $5M-8M. I was responsible for ideation, data acquisition, data base creation, feature engineering, proof of concept using out of time hold out data, and pitching the product to senior management. Expected release date: June, 2018. Tools used include Machine Learning, Spark, Hadoop, H2O, R, SAS, SQL, Python, Gradient Boosted Trees, Random Forest, Decision Trees, GLMs (Tweedie, Elastic Net, LASSO, Ridge)

Data Scientist in the Insurance Analytics division of Verisk Insurance Services. Primarily focused on creating, and then utilizing (modeling), new personal and commercial auto insurance R&D databases.

Lead researcher on three major projects where I designed, conducted, and analyzed output of experiments while earning my PhD. Major accomplishments include: 1) using large scale numerical simulations on a high performance computing cluster to discover and understand how different types of variation and assumptions on the nature of variation affect model predictions. 2) Using data extraction techniques to Scrape and Structure Unstructured empirical data sets that relate resource density to consumption rate. Then using model selection methodologies (AIC) I compared non-linear regression scores of the previously accepted model to novel models that I created and showed that the novel models outperform the previously accepted model. Additionally, the novel models are able to resolve two long standing controversies in theoretical biology. Tools used include Numerical Simulation, Distributed (Cloud) Computing, Python (PyCont, NumPy, SciPy), MatLab, R, Unix, SAS, Statistical Analyses (ANOVA, ANCOVA, Linear Regression, Non-Linear Regression, Akaike Information Criteria, Bayesian Information Criteria)

identifying novel questions, creating dynamical models, numerical simulation, statistical analyses, verbal and textual communication of findings

Provided instruction, clarification, and guidance for undergraduate students in both laboratory and lecture environments. Topics ranged from general biology to MATLAB.