LOS ANGELES, CA ***** •
SUMMARY OF QUALIFICATIONS
n Ph.D. Statistician with 30 years of research and consulting experience in Statistics & Probability. Major technical expertise in Construction of Predictive Models, Data Mining, Cluster and Segmentation Analysis, Pattern Recognition, Factor Analysis, Statistical Estimation Theory & Hypothesis Testing, Forecasting, Regression Analysis, Workforce Analytics, Marketing Analysis, Survey Design & Analysis, Statistical Inference from Samples of a Random Size, and Actuarial Mathematics & Risk Theory. Efficient at communicating with all levels of company executives. Distill complexity into salient points.
October 2016 – present Consulting LLC Owner and CEO Los Angeles, CA
• Provide consulting in Advanced Analytics employing my experience & expertise necessary for helping a company to be successful. Applying contemporary statistical & machine learning methods, which ensure that executive decisions are based on information extracted from data. It includes data visualization, data mining, building predictive statistical models, etc.
• Skills: Analytics Subject Matter Expert, Able to look at the business holistically, Provide efficient scalable solutions for business problems, Well versed at conveying complex analytical results to a non-analytical audience, Expert knowledge of Construction of Models, Data Mining, Cluster & Segmentation Analysis, Fraud/Pattern Recognition, Discriminant Analysis, Factor Analysis, Statistical Estimation Theory & Hypothesis Testing, Forecasting, Regression Analysis, Marketing Analysis, Survey Design & Analysis, Workforce Analytics, Statistical Inference from Samples of a Random Size, and Actuarial Mathematics & Risk Theory.
• Extensively use statistical/machine-learning tools (SAS Enterprise Guide, SAS Enterprise Miner, R, SPSS, Minitab, Statistica, KXEN, etc.) in extracting information from large datasets related to disparate sources. Am proficient in SQL, Access, Excel, Word, PowerPoint, and visualization tools such as Tableau. July 2013 – Oct 2016 Cisco Systems San Jose, CA
SME - Analyst
• Was developing predictive statistical models and conducting various types of data mining in a commercial environment to the end of solving diverse important business problems originating within Cisco Systems. Some of the successfully delivered projects are as follows:
a) Analyzed Return Materials Authorization (RMA) volatility in company’s products by using tolerance intervals and z-scores and have built a predictive statistical segmentation model that reveals RMA trend information by identifying and profiling segments with different RMA levels & ranks the involved predictors with respect to their importance.
b) With the objective to measure the impact of best practice (BP) rule violations on 2
syslog errors and, subsequently, to prioritize BP rule violations, conducted the analysis of relation between syslog errors and BP rule violations, identifying segments related to various rates of BP rule violations and have created the criterion for identifying a “bad”/”good” BP rule on a considered data set. c) Built robust and accurate predictive statistical models that allow evaluating banks’ network risk by utilizing software and hardware complexity & consistency metrics as predictors. The associated profile of the target variable describes each level of risk in terms of identified ranges of predictors.
d) Built a robust predictive logistic regression model for network disruption based on available information. The associated risk profile depicts pattern recognition in terms of identified ranges of predictors. This model is a building block for a warning system that will recognize symptoms of a serious error/disruption, which will occur in the next 10 seconds, and will allow taking an appropriate preventive action.
e) Proposed a design of Company’s Best Practices service request (SR) quality standards and outlined a Pattern Recognition method for identifying SRs that are open in a considered month & will be “incapable” of meeting these quality standards in the next month. The method is consistent with Statistical Quality Control Best Practices. The result is an Early Warning System, derived by building a predictive logistic regression model and by using the mentioned method for profiling the small group of recognized “incapable” SRs. f) Designed and implemented a holistic approach to Connected Analytics for Collaboration (CAC), which brings the CAC product quality to competitive standards that are required to make it useful for customers and distinguish it (with its unique advanced analytical features) from other existing Collaboration products. August 2012 – July 2013 DATABASE MANAGEMENT RESEARCH ANALYTICS, INC. Mountain View, CA
• Worked on a project for the client company RSA Medical. Applied statistical & analytical expertise to recommend and implement efficient health risk assessment. The analysis allowed significantly decreasing expected health care costs for individual members and/or member groups of individuals in a health care organization, and, as result, increasing profitability. October 2008 – July 2012 Toyota Financial Services Los Angeles, CA Customer Analytics & Modeling Manager
• Partnered closely with Marketing, Sales, Bank, Fraud, Audit & other teams in a fast moving environment and was using advanced research & statistical analysis to drive business insight & growth, as well as profitability and risk analysis.
• Acted as a bridge-builder across organizational boundaries with the necessary people and business groups.
• Constantly applied statistical & analytical expertise, especially strong knowledge & experience in the development of advanced statistical models and their application in a commercial environment, as well as in data mining and segmentation analysis. Designed and implemented the following models:
Ø Customer Value model
Ø Customer Loyalty model
Ø Price Optimization model for Internet Channel sales 3
Ø Response Model
Ø Bank Revolver Customer model
Ø PD/LGD segmentation models
Ø Sales Forecasting model
Ø Fraud related model
Ø Customer Service Inbound Calls Volume Forecasting model. 2006 – 2008 Countrywide Full Spectrum Lending Los Angeles, CA VP of Analytical Services
• Designed new models and methodologies allowing to evaluate /predict profitability and risk associated with the mortgage loans portfolio, as well as cost of marketing campaigns. Namely:
Ø Mortgage Early Pay-off model
Ø Mortgage Customer Retention model
Ø Mortgage Customer Delinquency model
Ø Delinquent Mortgage Customer Cure model
Ø Mortgage Customer Lead and Application models for various Media Channels Ø Customer Service Employee Retention model.
• Interfaced with business managers to obtain business requirements and presented results of modeling & analysis.
• Was responsible for customer segmentation analysis.
• Had direct oversight of the development, update, and maintenance of in-house models.
• Was responsible for month-to-month forecast. Evaluated data sources and analysis tools.
2004 – 2006 Marshall &Swift/Boeckh, LLC Los Angeles, CA Chief Statistician
• Was identifying and implementing efficient statistical methodologies necessary for making decisions based on large sets of data.
• Directed a group of statisticians in all upcoming projects, which included performing in-depth statistical analysis, extracting key information from databases, creating predictive statistical models (namely, Land Value model, Peril Ranking model, etc.), testing statistical hypotheses, identifying unusual behavior patterns in data, and other required statistical inference.
• Worked closely with other business areas (e.g., Marketing and IS) to brainstorm business issues, automate the process of statistical analysis, etc.
• Advised company employees on business related statistical & mathematical problems arising in their projects.
• Participated in pertinent discussions and/or meetings held in the company or elsewhere and was presenting findings to relevant professional groups/individuals. 2003 – 2004 The Tomás Rivera Policy Institute at the University of Southern California Los Angeles, CA
• Conducted statistical analysis of survey experience data by hypothesis testing, the 4
study of simple and main effects, by gaining a better understanding of the data prototypes that are in the study, by identifying the true pattern of relationships between variables, and by producing relevant numerical & graphical summaries.
• Utilized classical and advanced statistical methodologies, such as Kruskal-Wallis's test, Spearman’s rank correlations, Ordinal regression, Crosstabs, Chi-square analysis, OLAP cubes, Clustered Bar Graphs, histograms, etc. 2002 – 2003 Strategic Partners, Inc. Los Angeles, CA Consultant
• With the objective of driving decision making and new forecasting and inventory management processes (specifically, to optimize inventories), provided pertinent recommendations on the use of statistical forecasting techniques (time series analysis, etc.) necessary to validate and manage forecasting models of sales of apparel items manufactured by the company.
• Employed statistical & forecasting software (Minitab, Forecast Pro) to build and evaluate the identified forecasting models.
2001 – 2002 Los Angeles Unified School District Independent Analysis Unit Los Angeles, CA
• Utilized SPSS, HLM (hierarchical linear modeling statistical software), Access, and Excel (after screening, validating, cleaning the relevant experience data of the whole Los Angeles Unified School District, and merging data files with different key variables) to conduct statistical analysis, to build and evaluate regression models, and to design required tables and graphs.
• Based on the analysis results, provided the review of performance measurement instruments regarding student test outcomes and recommendations on use of outcome information in management of the District. 1999 – 2001 Netucation, LLC Los Angeles, CA
Director of Mathematics
• Was overseeing the work of other consultants.
• Designed mathematical models of various mathematical courses and subject- matter for pertinent databases.
• Provided strategic ideas for future mathematical projects.
• Using Quantitative Methods of Marketing Analysis & Data Mining, Survey Analysis, and appropriate Statistical tools, analyzed and evaluated survey reports prepared by a company specializing in marketing analysis.
• Conducted research to the end of identifying factors relevant to the future progress of Netucation.
1996 – 2000 Transamerica Occidental Life Co. Los Angeles, CA Consultant
• Developed statistical procedures for data analysis.
• Conducted statistical analysis of a Life Insurance Company experience data (after screening, validating, and cleaning large data files).
• Evaluated the risk of business by means of probabilistic comparison of the business models and designing a Reinsurance model.
• Developed and maintained the relevant relational databases, and, based on them, and Actuarial Mathematics methods, derived reports (retention & lapse analysis, pricing analysis) and recommendations about profitability of various life insurance products.
• Using Quantitative Methods of Marketing Analysis & Data Mining, and appropriate Statistical tools, analyzed and evaluated reports created by strategic marketing department to the end of identifying outliers and possible hidden trends.
1997 – 1997 Program for North American Mobility in Higher Education Los Angeles, CA
• Conducted statistical analysis of survey experience data.
• Produced relevant numerical and graphical summaries using Discriminant Analysis and various descriptive statistical methodologies.
• Based on the results of Data Mining & Factor Analysis, provided recommendations regarding selection of factors leading to the success of the Program.
EDUCATION & CERTIFICATION
1983 Moscow, USSR
Senior Research Fellow in Probability Theory and Mathematical Statistics 1975 Steklov Mathematical Institute Leningrad Branch, St. Petersburg, USSR Ph.D. in Mathematical Statistics
1971 Tbilisi University Tbilisi, USSR
MS in Mathematics
1996-1998 PASSED ACTUARIAL EXAMS 100, 110, 120, 140, 151 SOFTWARE PROFICIENCY
SAS Enterprise Guide, SAS Enterprise Miner, R, SPSS, Minitab, KXEN, Statistica, Insightful Miner, Clementine, Model Builder, HLM, Forecast Pro, SQL, Tableau, Excel, Word, etc.
3 MONOGRAPHS, 44 SCIENTIFIC PAPERS, 25 RESEARCH ANNOUNCEMENTS
(THE LIST OF PUBLICATIONS AVAILABLE UPON REQUEST). 6
References are available upon request.