Sam Whitehill Email & Phone Number

Los Angeles, Ca, Us

Virtual

Developing solutions to short term contractual projects on one of the world’s largest freelancing websites. Established a long term track record of 5-star ratings from clients on over ten projects.● Currently reducing costs for a logistics technology company by developing a Python API to optimize supply locations for maximum demand fulfillment.

Woodland Hills, Ca

Developed machine learning models and performed data analysis to benefit Underwriting and Product management, across a variety of use cases. Created Tableau dashboards, data visualizations and presentations to make recommendations to senior stakeholders. Gained deep expertise across the Business Insurance Group, by researching and solving diverse business problems.• Proposed & built a groundbreaking CNN image classifier with a projected ROI of 65%, to identify buildings in poor condition. Created a labeled image dataset of buildings in poor condition by researching Underwriter journals.• Proposed & created a competitive benchmark model via building a CNN image classifier (88% accuracy) to find swimming pools & tennis courts, based on a manually labeled image dataset.• Reduced premium leakage by developing a machine learning classifier to identify under priced policies. Improved model accuracy by 5% via manually validating & cleaning ground truth labels.• Increased premium revenue on the Real Estate portfolio by 6%, via building a fuzzy address matching application to identify strip malls misclassified as office buildings.• Improved profitability & yield rates (quotes) on the retail, office, & service portfolio by co-developing a GLM pricing model. Automated feature engineering & selection (via Python) to choose the top 5 features from an initial set of 2,000.• Increased marketing leads 5% by building a machine learning classifier to predict the SIC from vendor data.• Established a proof-of-concept Natural Language Processing (NLP) model to identify Underwriting comments most predictive of future claims.• Identified key reasons for lost customers by developing machine learning topic (NLP) model to parse customer survey comments. • Prevented a flawed retention model from going live by performing a complex peer review to find target leakage.• Assisted actuaries in reducing wind & hail claims by creating a clustering model from complex FEMA datasets.

New York, Ny

Developed financial models using Python and MATLAB. Programmed MATLAB applications to back test an Interest rate forecasting model developed by the American Academy of Actuaries.•Built a Maximum Likelihood Estimation program in Python (including NumPy and SciPy libraries) to estimate the parameters of an Interest rate model.•Implemented multiprocessing functionality in Python, to obtain better runtime efficiency.•Constructed an optimization model in C# using Delegates and Multithreading as well as a performance metrics application in Java.•Supported an end user insurance model for 150 global users.•Performed automated quality assurance testing of financial ratings and scoring models via Python and MATLAB. Significantly reduced critical software defects in production credit rating models.

New York, New York, United States

• Proprietary trading of stocks using technical trading strategies.

New York, New York, United States

Developed credit derivative pricing models in C#, and in Excel (VBA). Covered products such as ABS, CDOs, Credit Indices, RMBS, and Single Name CDS.•Built a Synthetic CDO pricing model using Quasi Monte Carlo methodology. Model used a low discrepancy sequence (Halton sequence) as well as a correlation matrix.•Created Value at Risk reports and Economic Risk Capital reports for Cash & Synthetic CDO trading desks.

Charlottesville, Virginia, United States

Built n-tier enterprise systems using tools such as Visual Basic, VBA, SQL Server, and Crystal Reports.•Constructed an online tickets application using ASP.NET, XML and XSL. Created user interfaces, wizards, and COM business components using Visual Basic.•Created statistical reports for senior management via VB, ADO, stored procedures.

Charlottesville, Virginia, United States

Built a wide variety n-tier applications consisting of user interfaces, COM objects, and stored procedures using Visual Basic and SQL Server.•Created reports and graphic exhibits for hospital management using complex SQL queries, Excel spreadsheets, and Crystal Reports.•Constructed sophisticated Active Server Pages using HTML, VB script and JavaScript.

Nanodegree, Machine Learning Engineering

Activities and Societies: Building Machine Learning projects using Python and associated libraries. Learn about state of the art machine.

Bachelor Of Science, Commerce; Finance And Management Information Systems, 3.6 Gpa

Graduated with Distinction (awarded to top 20%).