Dr. Thomas A. Woolman

• For a variety of strategic national security projects including:
• Deep learning neural network models for risk scoring of qualitative international entity behaviors using autoencoding models
• Neural embedded natural language processing for classification predictive models related to continuous monitoring of targeted international entities
• Using ensemble neural network models with ranked voting to minimize Type I and Type II statistical errors in predictive models while conducting ensemble learning for feature reduction
• Incorporating weighted graph models for calculating multi-level relationships between international entities comprised of successive relationship layers
• Deep learning and neural-embedded models for natural language processing predictive models
• Developed a deep learning predictive model for qualitative response behaviors that predicted causal relationships in human behaviors along with a permuted test of statistical significance for these models
• Created a series of deep learning autoencoding predictive models for unsupervised feature reduction and anomaly detection related to malware behaviors on national security computer networks
• Developed a test of statistical equivalence for this model to validate performance based on a baseline validation model
• Deep learning self-supervised and supervised financial fraud detection models with Platt probability scaling and permuted tests of statistical significance
• Developed a deep learning predictive model for calculating enterprise risk scores and risk causality factors for national security capital project risk scoring
• Utilized Explainable AI models and methods
• Deep expertise with ensemble deep learning time series predictive models
• Expertise in utilizing nonparametric permutational statistical tests for addressing research questions and null hypothesis tests from non-linear AI/ML models., Lead data scientist for a cyber analytics security project that utilizes nonparametric statistics as well as supervised deep learning and self-supervised deep learning models.

• Data scientist and systems architect for a Federal law enforcement agency project
• Top Secret security clearance held
• Developed supervised and unsupervised / self-supervised deep learing predictive models and model ensembles
• Utilized tests of statistical equivalency and permuted tests of statistical significance for proving statistically significant model efficacy
• Created mixed methods research predictive models using qualitiative and quantitative datasets
• Developed predictive (regression and classification) models using R and RStudio in an Amazon Cloud (AWS) environment, using SQL data sources integrated into R
• Developed Explainable AI models to support detailed “plain English” model prediction interpretations and narratives, along with producing visualizations to support the detailed explanations.

• Lead data scientist responsible for design and implementation of advanced statistical modeling features
• These include the use of Python and SAS Enterprise Miner for predictive modeling, machine learning and data mining functionality along with standard normal distribution scorings, supervised and unsupervised machine learning models, ANOVA analysis functions and chi-square validations across disparate size datasets within the overall system design
• Lead technical author for the company’s patent filing, “SELECTION SYSTEM FOR MACHINE LEARNING MODULE FOR DETERMINING TARGET METRICS FOR EVALUATION OF HEALTH CARE PROCEDURES AND PROVIDERS”, filed with US Patent Office #075556-001100PL02
• Patent granted November 2023, U.S
• Patent number 11,816,539
• Responsible for technology management of SurgeonCheck's analytical Business Intelligence products focused on surgeon performance by both qualitative and cost quantitative metrics across hospital and insurance networks
• Lead developer for business intelligence cloud-based analytics package framework
• Responsible for annual IT and software application budget development and project implementation scheduling

• Data scientist and systems architect for an advanced artificial intelligence and analytics project involving R and Deep Learning, Quantitative Statistical Modeling, Multivariate Classification and Multiple Regression, Machine Learning predictive modeling and anomaly detection.

Teaching primarily graduate and some undergraduate data science and analytics courses, including:

• DAT-650 Advanced Analytics
• DAT-640 Predictive Analytics
• DAT-610 Optimization and Risk Assessment
• DAT-520 Decision Methods and Modeling
• DAT-510 Foundations of Data

• Data scientist and systems architect for an advanced artificial intelligence and analytics project involving R, Python, Quantitative Statistical Modeling, Big Data (Databricks, Spark, MapReduce), unsupervised classification, supervised predictive modeling and multivariate cluster modeling with quantitative explainable AI methods, plus time series predictive modeling.

• Scientist and systems architect for a high-level Federal agency planning, budgeting and capital execution project.

• Financial analytics and data mining system

• Hyperion Planning architect responsible for redesign / optimization and new development to support new corporate planning and forecasting initiative
• Developed Planning applications to support finance (B/S and P&L), sales and HR planning and forecasting as well as international sales flash applications
• Developed calculation scripts and Hyperion Calculation Manager rules to allow integration between sales and finance plan types, produced exponential smoothing and linear regression forecast models and supported what-if analysis across multiple plan types
• Led development and enhancements of a sales operations metrics application in Hyperion Planning that calculated sales cost metrics across all product and raw materials combinations for all domestic and international business units
• Developed reporting application to allow for the analysis of these cost metrics
• Developed a sales revenue and inventory linear and nonlinear regression model within Essbase to seed the monthly forecast cycle with data that captured seasonality trends and conduct pattern discovery across multiple geographic regions and currencies using time series data
• This was utilized as the basis for a customized demand planning solution using Python
• Developed Essbase ASO cubes to support Planning analytical reporting across all plan types and statistical forecast modeling of details sales trends down to SKU levels across all entities
• Led prototype analysis and feature comparison review between IBM Cognos BI and Oracle Hyperion
• Prototype metadata management solution with Hyperion DRM
• Worked extensively with FDMEE to facilitate data mapping and fact data loads into Hyperion Planning and Hyperion Essbase from multiple G/L source systems
• Led development of a daily sales flash reporting application that utilized linear regression, double and triple exponential smoothing and variable weighted averaging calculation scripts
• Integrated SAS Forecast Server datasets into Essbase that created a series of hybrid forecast models for analysis and reporting in the cube
• Supported 24x7 management operational reporting out of Essbase utilizing custom built Excel-based reporting package that extracted data from multiple Hyperion Planning plan types as well as Hyperion Financial Management (HFM)
• Travelled to China and Thailand to support local corporate divisions with planning and forecasting needs and customizations
• Designed and implemented Hyperion Planning to HFM plan and forecast data integration.

• Financial analytics and data mining system

• Designed and implemented company’s restarted Biogeochemistry Prospecting Project focusing on base metals, rare earth elements, and platinum group elements in an area encompassing the Mesoproterozoic Karagwe-Ankolean sequence within the Kibaran Fold Belt of NW Tanzania
• Lead statistical modeler and systems architect for an innovative, advanced analytics project utilizing state-of-the-art forecasting methodologies and biogeochemical data gathering applied to an extensive mathematical geology database
• Published scientific author on the field of advanced analytics and business intelligence based on groundbreaking work conducted with this client
• The BGC Survey searched for base metals, rare earth elements, nickel, platinum group metals suite and cobalt target elements in addition to their known geochemical indicators (pathfinders), including Cr, Cu, Zn, etc
• A multidimensional correlation matrix with dimensionality factors based on plant species, seasonality, soil and target elements was used to optimize the response ratio metrics upon analysis of lab sample results before splining the calculated BGC response ratio data to the geophysical map
• Factor analysis, BAC coefficients, calculated response ratios and statistical kriging was utilized for value inference and anomaly detection of target and indicator elements which allowed us to conduct online analytical database mining (OLAM) operations within Essbase and R Studio
• Utilized data mining concepts including decision trees, factor and correlation analysis models, ANOVA analysis of variance, neural nets, statistical response ratios, nonlinear regression forecasts and the development of an Essbase data mining framework model to successfully analyze and forecast biogeochemistry data and discover precious metals and base metal mineral deposits and assist in determining which property exploration licenses to keep and which licenses to drop
• Frequently interacted with senior geologists, biochemists, and geophysics team members to lead the statistical data analysis and development of software systems to generate these forecasts
• Made use of Hyperion Essbase ASO cubes, R Studio, Python, SAS Jmp Pro, Tableau, Tiberius Data Mining, SAS Enterprise Miner, and Micromine software applications to conduct this system development
• Presented to the shareholders and board members as a keynote speaker at the company’s annual general meetings
• Managed a full-time team of approximately 12 personnel for field operations including botanists, geologists, geotechnicians and lab technicians while directing exploration target selection and statistical data analysis, while managing logistical planning and budget planning
• Frequently interacting with third party lab technicians as well who supply our 43-101 compliant laboratory results data.

• System architect for Hyperion Planning EPM System 11.1.2.2 corporate enterprise planning initiative
• Designed and managed Hyperion DRM integration and application development to process enterprise standard dimension metadata from multiple source ERP systems into a centralized change management repository
• Designed Hyperion DRM dimension metadata extracts
• Primary technical resource within a diverse team of analysts and developers which designed a comprehensive enterprise planning solution for all of company’s worldwide segments and divisions
• The design integrated dozens of systems including legacy Hyperion applications into a robust, unified enterprise planning solution
• Managed corporate finance analytics review, data gathering and systems analysis functions across geographically and functionally diverse segments of the corporate enterprise and successfully designed a solution which integrated all of their diverse needs and process calendars into a single integrated system
• Provided data governance and management planning and control across a single commercial-view of their multifaceted business environment
• Designed FDM and FDMEE data load and transformation processes to support ERP fact data loads to Hyperion Planning
• Designed and led technical implementation of Hyperion Planning Enterprise Balance Sheet, Enterprise Cash Flow, Enterprise Revenue and Enterprise Expense applications to support multi-division corporate planning and AOP cycles.

• And senior developer for Hyperion Planning EPM System implementation for Exelon, subcontracted through Deloitte Consulting
• Developed a new Nuclear Power model for Hyperion Planning to allow end users to enter budget and forecasting financials and labor data, and conduct analysis of the consolidated results
• Consisted of multiple Plan Type cubes with several dozen business rules and approximately 30 calculation scripts and a number of complex Hyperion Extended Spreadsheet Macro workbooks
• Prototyped Hyperion DRM application to integrate master metadata management for major dynamic dimensions in Hyperion EPM
• Replaced IBM Cognos BI Report Studio and Analysis Studio legacy functionality with Hyperion Essbase, Hyperion Financial Reporting and Hyperion Planning applications
• Designed and implemented highly scaled corporate headcount analytics ASO cubes to support Federal Energy Regulatory Commission reporting applications, including MDX scripting to support waterfall allocations within the ASO cube
• Made extensive use of Unix korn shell scripting and Maxl scripting language to automate a wide range of Oracle Hyperion processes
• Managed a team of Hyperion developers including myself and two other Essbase/Planning developers plus off-shore resources
• Made use of Hyperion Shared Services security model to manage LDAP user IDs to create provisioned security groups to allow different aspects of the enterprise to utilize the same application for budgeting, planning and forecasting and view different sets of metadata and fact data within the same application
• Designed and implemented an Essbase allocations model to support Baltimore Gas and Electric data from Primavera systems, utilizing a range of Essbase BSO and ASO calculation scripts
• Developed budgeting, planning, forecasting and analytical models to support multiple segments of the company including nuclear power, Constellation Energy, Philadelphia Electric Company, Exelon/ComEdison, Baltimore Gas and Electric, and non-generation corporate operations
• Worked in an intense U.S
• FERC-regulated nuclear power 24x7 managed environment that maintained a high-availability environment to provide logistical analysis planning, labor analytics, budgeting and forecasting systems and reporting databases.

• And senior developer for Hyperion Planning EPM System 11.1.2.1 implementation for Southwestern Energy, subcontracted through Deloitte Consulting
• Led a team of 4 developers in addition to myself to redesign and solve complex Hyperion Planning, Hyperion DRM master metadata management and Essbase calculations and optimizations including salary and headcount planning, revenue, expenses, corporate, midstream and exploration & production capital, hedging and seasonal/area differentials, risk modeling and equity/cash flow management
• Led the redesign and developed extensive set of calculations for a dedicated Economics statistical model using Python and Essbase which calculated derived data from Halliburton’s ARIES Reserves Management Database System via a customized staging table solution
• The Essbase Economics cube allowed for the analysis of oil and gas volumes, development and exploration expenditures, and statistical KPIs down to the well, area, field or subsidiary company, for actuals, budget or forecast across multiple years in far more granular detail than supported by the Hyperion Planning system
• Lead developer for an intricate ASO Essbase Allocations Engine application which accepted G/L feeds from Oracle into a 14-dimension Essbase cube and generated debit and credit-side allocations across multiple subsidiary companies by accounts and budget centers, for actuals, budget and forecasts, parsed and then uploaded back into the Oracle G/L
• Budget and Forecast allocations were also then fed back into Hyperion Planning for use in the company’s planning and forecasting processes
• Created dozens of large statistical models as well as planning system integrations, statistical formulas and python scripts based on legacy IBM Cognos Analysis Studio applications
• Worked closely with client senior financial management end users to design dimensions, data flow processes, data forms and business rules to support the budget planning cycle.