Director Analytics

JB Ph.D

 

Profile

 

• Proven self starter in quantitative modeling with hands on programming, mathematics and research skills

• US citizen, flexible for travel and relocation (renter / non-homeowner), living in Washington, DC

• 15+ years of statistical modeling with SAS in industry and academic settings on PC and UNIX platforms

SAS: BASE, Macro language, SQL, STAT, GRAPH, IML (Matrixes), EM (Enterprise Miner / Data Mining)

• ANOVA (PROC GLM), Regression (PROC REG), Multinomial logistic regression (PROC LOGISTIC, PROC CATMOD), Survival/Hazard Models (PROC PHREG), Canonical Correlation (PROC CANCORR)

• Data Mining, CHAID / CART decision trees (SAS Enterprise Miner and ANGOSS Knowledge Studio)

• Extensive programming in Microsoft Excel and Microsoft Access using Visual Basic for Applications (VBA/macros)

Monte Carlo simulation modeling using SAS, Excel VBA (macros), Crystal Ball, @Risk, Extend & AutoMod

C and C++ for developing heuristics in Ph.D. program; completed college classes in C++ and Visual Basic (VB)

• R Programming language (Time Series, Predictive Modeling / Machine Learning)

 

Education

 

Ph.D.

Purdue University

Krannert School of Management

Management Science / Quantitative Methods Area (“Operations Research”)

Coursework consisted of Statistics, Industrial Engineering, and Research Methods

Research consisted of Monte Carlo Simulations and Data Mining / predictive modeling in SAS

Recipient of Bilsland Fellowship and Frederick N. Andrews Fellowship

  Dec 2007

MBA

Purdue University

Krannert School of Management

Management Information Systems Concentration

May 2001

BS

University of Richmond

Robins School of Business

Business Major; Finance and Marketing Concentrations

May 1996

 

Professional Industry Experience

 

Marketing firm, Director of Analytics                                                     Apr 2013-Jun 2013

Recruited to a high tech startup by co-founder of a former employer

Work with Big Data (billions of records of data), Amazon Cloud S3, Google Big Query, Hive/SQL, SQL/Dynamic SQL

Extensive use of R Programming language

Build predictive models (linear models, logistic regression, recursive partitioning, variable selection, Ridge regression)

Financial Firm, Principal                                                                                       Nov 2012-Apr 2013

Hired to assist on Independent Foreclosure Review projects, which ended January 2013

Exposure to banking and regulatory issues, such as Stress testing (PD, LGD, EAD) & Anti-Money Laundering compliance

Preparation work for Fair Lending Review and analysis of Disparate Impact of Lending practices

Extensive use of R programming language for Time Series and Predictive Models (logistic regression, machine learning, support vector, decision trees/ recursive partitioning)

 

Financial Firm Vice President, Risk Assessment                                                                                                                      Feb 2010–Nov 2012

• Hired by C-suite of financial services firm to build (credit) risk assessment modeling, analytics, & reporting tools

• Set up and configured servers (Microsoft Windows Server 2003 R2)

• Purchased and installed Microsoft SQL Server 2008, SAS 9.2 and SAS Enterprise Miner 6.2

• Acquired data feed from loan management system vendors (daily FTPs)

• Acquired data from subprime credit reporting agencies (Clarity, Experian Equifax, Lexis Nexis, Data X, Targus Info)

• Wrote all SAS, SQL, Excel VBA (macros) code from scratch (50,000+ lines of code)

• Implemented fully automated overnight data reporting and error handling which FTPs and/or emails 100s of fully formatted Excel reports and event alerts to Operations, Treasury, Accounting, Marketing, Call Center, etc.

• Work with data vendors to back test loan performance with alternative data sources to improve loan underwriting

• Built multivariate statistical scoring models (Regression, Logistic, and Canonical Correlation) and Decision Trees (CHAID / CART) to determine underwriting risk. Wrote macros to create decision trees without Enterprise Miner

• Responsible for adjusting underwriting filters for online loan approvals using insights from scoring models

• Developed algorithm for loan credit line increases, fully automated delivery to operations, providing AR growth

• Analysis, trending & projections used to improve profitability & assist in securing investment capital for growth

• Recruited and hired 2 team members (1 full time employee and 1 intern who became a full time employee)

                                                                                                                                                                 

DOTOMI, Inc., Analytics, Director                                                                                    Oct 2007-Feb 2010

• Early hire at a rapid growth advertiser which was eventually acquired by ValueClick for $295 million

• Architect of the yield management algorithms for rapidly growing online advertiser serving 500 million ads monthly; prototyped system then worked with technology team to implement and ensure accuracy; Developed models for estimating the value of each user’s online advertisement.  Worked with technology team to read in each client’s specific model (Log Linear Regression models in SAS) and then use those models to drive the message prioritization algorithms, media spend and forthcoming real time bidding. 

• Monte Carlo simulations for back testing 100s of millions of data points for optimizing revenue across clients

• Resulting models, calculations and decay curves are in production on server;

      results achieved 6-8% top line revenue increase above previously random assignment.

• Developed SAS statistical models and analytical tools for marketing effectiveness and profitability

• Maintain system of SAS programs that run each night to provide analytic reports for the start of each day

• Independently identified bottlenecks in media buying process that were constraining revenue opportunity

• Conducted numerous statistical tests with random splits of users; results identified opportunities for revenue gains, cost savings, as well as discovered additional cost drivers

• Developed and maintained automated Excel VBA macro tools used daily by all account managers used as a decision support tool to monitor monthly trends, ROI goals and financial projections (forecasts) for each client

• Applied matrix calculations (SAS/IML) to determine overlap of media networks to assess unique value of each

• Pioneered new Sales Planning Tool (Excel VBA Macros) that incorporates decay curves to account for diminishing marginal utility (i.e. non linear projections).  Tool automatically determines the parameters of different deal structures that will generate equivalent revenue for a company to assist in negotiating deals.

• Created and implemented logistic regression models for scoring panel of profiled users to target for advertising campaigns using demographic attributes from 100+ million households (SAS/STAT, PROC LOGISTIC)

• AWARDS: Two time recipient of “Chairmen’s Reserve” award, consisting of stock grant and public recognition from the CEO at quarterly company town hall meeting recognizing “outstanding contributions” to the company.

 

International Business Machines (IBM), IBM Global Services, Associate Consultant (IBM employee) 2001- 2002

• Served as project manager on-site at New York DOT developing process improvements for their billing system

• Developed automated spreadsheets in Excel with VBA macros to improve accuracy and efficiency

 

Purdue University, Graduate Assistant & Senior Webmaster in the MBA Career Services Office     1999- 2001

• Extensive programming with Active Server Pages (ASP), HTML, VBScript, JavaScript, SQL, PHP, MySQL

• Developed web site error handling and reporting for quality improvement initiative

• Converted many static web pages into reliable online applications

• Database programming and design in MS Access with VBA (macros) and SQL

 

Lucent Technologies, Subcontractor (not a Lucent employee)                                                  Summer 1999           

Independent consultant for self sold project to automate quarterly tax estimates

• Developed automation routines and user interface in Excel VBA for accounting department

Allowed one manager to complete a process previously allocated to 6 people

• Calculated each state’s tax liabilities based on various State Income Tax Filing Methodologies: Separate, Combined, Domestic Combined, Worldwide Combined, Water’s Edge Combined, Nexus Consolidated and Nexus Combined

 

Information Resources Inc. (IRI) - Analytic Insights Group, Project Director / Analyst                                               1996 - 1999

• Worked in the advanced analytics / consumer insights modeling group of a large marketing research firm

Prepared and delivered marketing mix models to firms in the Consumer Packaged Goods Industry

Enhanced and maintained departmental SAS statistical models to conduct price and promotion analysis (SAS/STAT)

• Developed quantitative decision support tools to simulate market results and select optimal product mix (Excel VBA)

• Created departmental technical training manual and numerous automated programs for increased accuracy & efficiency

Recipient of four “extra mile” awards; award approved by Vice-President upon peer nomination

  Contact mary@analyticsearches.com.