SUMMARY:

• Results - oriented professional with diverse background offering senior-level experience in the areas of data mining, risk management and statistical modeling.
• 15 years of experience creating advanced quantitative and predictive models using SAS and SQL.
• Lead efforts that resulted in creation of Behavioral, Acquisition, Churn, Fraud, and B2B models to segment customers, reduce bad debt, and balance the company’s risk portfolio.
• Enterprise Subject Matter Expert for the development, design, and testing of predictive models for credit scoring systems and behavioral marketing analysis.
• Enjoys influencing the corporate strategy position by delivering fresh statistical solutions.

TECHNICAL SKILLS:

SAS: SAS Enterprise Miner, SAS Enterprise Guide; SAS ETS, SAS Programming

Database Tools: SQL Server, Teradata, Oracle, Business Objects, Access, MS Excel, Tableau

Operating Systems and Development Platform: Windows, Unix, Visual C++, R, VBA

Data-mining Methods: Decision trees, Linear regression, Logistical regression, Neural Networks Clustering, Spectral Analysis, ARIMA, Survival Analysis, Text Mining (Sentiment Analysis), Gradient boosting

Data Science Tools SAS, TIBCO Statistica, R, Python, C#, CNTK Deep learning, TensorFlow, TLC, Spotfire

Web Tools: HTML, OMNITURE, XML

PROFESSIONAL EXPERIENCE:

Confidential

Data Scientist

Responsibilities:

• Used Gradient boosting machine learning techniques and CNTK (python) deep learning algorithms to identify bad actors committing fraud and piracy schemes.
• Developed unsupervised learning model for targeting clusters of customer behavior among the MPN population using R and SQL.
• Improved the KMS investigation process by developing a XG Boosted Tree model to segment and identify potential software piracy populations.
• Improve the clients VLML model by 12% using TLC SVM machine learning platform to construct a new model for screening this segment of customers.
• Designed a risk tool that segregated good actors from bad actors using R survival analysis to reveal their activation and other stratified patterns.

Confidential

Statistical Consultant

Responsibilities:

• Leader of an international team of statisticians that maintains a SAS based analytical data-mart, design experiments and review analysis to support the EDC CRM group. The team has provided analysis of: tenure persistency; credit portfolio trending; and customer demographic analysis.
• Presented and built a SAS cluster model solution to identify consumer segments that drive performance for a Fortune 50 bank’s portfolio. This product’s implementations lead to an ongoing revenue stream for Confidential and the client.
• Produced a survival analysis study that explained the relationship between alerts and tenure persistency for the Free Credit Report consumer base.
• Responsible for the Confidential Consumer Direct (ECD) Plus Score. The Plus Score is the primary product for ECD’s subsidiary company Free Credit Report.com.
• Responsible for Competitive Analysis Report (CAR). CAR is used for tracking customer churn and for reporting the company’s market share. ECD C level executives presented the CAR report before the Justice Department and FTC when making the case to acquiring new companies.
• Responsible for National Score Index (NSI). NSI is a web based report that provides a snap shot of U.S. consumer credit behavior and the country’s financial health. The NSI‘s data has lead to new products acquisitions with B2B clients.

Confidential

Senior Manager of Analytics and Modeling

Responsibilities:

• Served as a corporate subject matter expert for statistics and data mining. Implemented methodologies for evaluating the effectiveness of targeted campaigns and other business changes.
• Data Mining: Developed a logistic regression model and strategy to segment and rank customers based on their probability to convert to paying subscriber status. This model was the basis for most A/B treatments and is used by the C level executive staff to monitor and plan the Classmates roadmap.
• Introduced auto regression models for determining the value of user generated content on client’s site.
• Used survival analysis methods to develop time to event models for predicting time until a customer converts to a paying subscriber status and also to predict subscribers churn.
• Developed decision trees to describe factors most effective for designing profit-driving treatments.

Confidential

Manager, Risk Management / Data mining

Responsibilities:

• Introduced SAS as an enterprise analysis tool within a week of joining the company. Built a data-mining center by establishing an analytical data mart and upgrading to SAS Enterprise Miner.
• Risk Management: Enterprise Subject Matter Expert for the development, design, and testing of predictive models, credit scoring systems, and SAS code applications. This effort has resulted in Behavioral, Churn, Fraud, and B2B models and lead to customer segmentation and reduction of bad debt.
• Data Integration: Using a SAS analytical data mart and data mining, implemented predictive models that led to actionable results and aided company in focusing on the most profitable customer segmentations and created SAS scorecard-monitoring utilities.
• Manager: Conceptualized, developed, and managed a department, which was assigned to control corporate write-off, churn and introduce new technologies for the Marketing and Sales departments. During the last year of my tenure, the prediction models and segmentation work done by my team and myself resulted in a 43% reduction in bad debt and a 26% reduction in involuntary churn.
• Interacted with VPs on a regular basis for such goals as keeping them updated on major topics of concern (churn reduction, fraud, corporate write off and joint strategy sessions for similar results.)
• Monitored and validated models and data from Confidential, Trans Union, Equifax, and Fair Isaac.