-Provided data science consulting with several startups (of which Greenlight invested) on a myriad of insightful projects from pricing, conversion rates, severity, timeliness of deliveries, as well as road mapping data for future utilizing cutting edge algorithms. -Created predictive models for mortgage default rates, homeowners and automobiles pricing using gradient boosting and GLM-Developed a centralized risk management database that combines initial expectations with actual development -Developed new methodologies and process improvements around reporting requirements including BCAR-Analyzed and integrated catastrophe modeling results for pricing, cat management thresholds, and regulatory reporting.

Risk Modeling• Created new features in fraud models that increased the fraud capture rate by as much as 100% in certain segments using Hadoop• Trained colleagues and built prototypes in hadoop on a spectrum of applications• Spearheaded inclusion of 3rd party data into various existing fraud and authentication models• Built a statistical model for credit card “testing” for earlier fraud detection in HadoopMarketing• Built predictive statistical models using a variety of techniques to optimize prospect list for marketing campaigns• Assessed results of co-branded marketing signup campaigns• Develop synthetic cohorts to use as a control group to assess the impact of promotion activities, using both individual and aggregate controls.• Create Factor analysis / Principal Component Analysis for various uses• Create applications in Hadoop to rapidly synthesize disparate information

• Developed compromised-site monitoring and detection algorithms by creating simulated data to train and refine algorithms • Created a production quality process incorporating Patient Rule Induction Method (PRIM) to generate authorization accept/reject rules for issuing banks– received a U.S. Patent.• Produced a process to create a question/answer series for cardholder authentication system based on transaction history

Responsible for developing statistical models to quantify loan rates and delinquencies• Created the company’s first logistic model of default frequency

Responsible for creating corporate-wide production systems to monitor and quickly understand financial results. • Designed and built a critical financial monitoring tool integrating the most recent actuarial analysis with real-time claims data. This suite of tools quickly became one of the most important financial monitoring.• Conducted logistic modeling for all personal lines for accurately pricing strategies• Redesigned and implemented an application to allocate $3B in reserves including adding audit capabilities, enabling new methodology for allocating reserves and annual savings of $180K in processing costs• Served as single point of contact between managers and IT engineers to translate business requirements into technical specifications

Managed team to model and analyze data related to insurance pricing and risk management including reporting to senior management and managing client relationships• Promoted to Assistant Vice President, managing a four person team, from Actuarial Student through various roles over the course of six years• Analyzed and set reinsurance pricing structures using increased limits factors analysis, catastrophe modeling, post-modeling simulation and other analyses tools• Conducted feasibility analyses and audits, developed detailed pro-forma financial statements models, evaluated buyout opportunities, and used both traditional methods and software to manage loss reserving• Designed and launched a data warehouse on a UNIX platform in SAS, resulting in annual savings of $200K and created a program to enable underwriters to run reports from their desk• Forecast loss reserves for all of warranty business – automobile and appliances