Designed, developed, and maintained a robust data infrastructure with automatic anomaly detection and reporting, enabling seamless analysis, machine learning, and AI tasks.Built a comprehensive machine learning infrastructure in Azure and GCP, empowering NLP / LLM, Computer Vision, Deep Learning, and other Data Science tasks.Implemented a CI/CD environment using GitHub to streamline development and deployment processes.Authored comprehensive unit test cases for all functions and routines, utilizing automated testing in Git to ensure code quality and reliability.Developed insightful dashboards in Power BI, leveraging Azure data storage to provide actionable insights to management and clients.Created detailed technical documents outlining the construction and maintenance of all infrastructure, models, functions, and workflows for easy reference and knowledge sharing.Collaborated effectively across departments and with clients to identify, frame, and provide valuable analysis and insights to address challenges.

Stellar Grove is a small company performing several different services. While we specialize in data, AI/ML and Applied Mathematics, we'd be remiss if we didn't at least mention our design and production capabilities. At Stellar Grove we look to help unlock the wonders of life using whatever tools we have at our disposal.

• Automated creation of Financial Statements and other accounting functions using Python and VBA.• Provided Management & Accounting Consultation to mid-size industrial clients for improved financial strategies.• Designed and constructed Account Reconciliation routines to streamline processes and enhance accuracy.

Built and deployed a product price registry that allowed us to increase our coverage of COGS for potential borrowers by an average of 20% across all covered markets.Derived, Constructed and Implemented KPI’s into production code base to advance the proprietary Underwriting Algorithm using a CI/CD approach to changes in production. KPI's were used to score current and potential borrowers based on credit worthiness in an automated fashion.Used statistical tests and measures of central tendency to determine confidence intervals of each portion of the KPI being built.Built ad-hoc functions like: determining the angle of a sales trendline based on 18 months of historical data, conversion functions and other data transformation / processing functions.Authored unit test cases for all functions and routines added to the automated underwriting algorithm as part of the production code deployment workflow.Contributor to Data Engineering production code base implementing changes to Google Cloud Platform databases using DBT and git.Constructed Mode BI Dashboards to monitor KPIs introduced. The dashboards would incorporate current and past values and highlight anomalies using visualizations like Statistical Control Charts and other outlier visualizations.

I served as the Chief Financial Officer & Chief Technology Officers.As CTO, IImplemented ML algorithms to multiple classification processes that were originally done manually, saving time, money and reducing errors in the process.Built and deployed birth / death models, linear regression and time series models to forecast patient count, average spend and top line sales figures and expense figures. Modeled forecasts were within 2% of actual sales each year and within 5% on expenses.Built and implemented ensemble models to determine likely basket of goods for patients based on prior purchases.Used ensemble methods (NLP, logistic regression and unsupervised learning) to analyze purchase data, online reviews and other external data to segment our patient base for targeted marketing.Engineered and implemented all data and ML infrastructure in Microsoft Azure including automated ETL processes with error logging and exception reporting.Created a scoring algorithm for our vendors based on sales velocities of products, profit margins, time to deliver and relationships with our staff.As CFO I:Formulated budgets and technology buildout for our initial and upgraded brick and mortar retail location.Derived and monitored monthly budgets of over $350 for our departments.Created an automated variance analysis process that would look at actual versus estimated spending based on transactions from multiple data sources.Created Power BI dashboards to monitor all data created throughout the automated process environment. Dashboards would update hourly and were available to various levels of the company with varying levels of security.Grew discounted sales 240% from 2018 - 2022, resulting in multiple above market acquisition offers.Used a combination of EBITDA multiples and discounted cash flow (DCF) models to set the sales price of the company. Discount rates were derived by mining data to create a risk profile of companies that operated in alternative industries.

Created predictive algorithms using Python (ANN, Regression Forest, Decision Trees) for use across the health system. Models included predicting patient discharge disposition, likelihood of readmission, likelihood of cancelled / no show appointments and predicting hourly census amongst others.Created and tracked the performance of several NLP models, to perform sentiment analysis to drive higher patient satisfaction scores.Data Lead for the Opioid Task Force put together by AAHS Leadership to combat the opioid epidemic in our surrounding area. Partnered with clinicians to develop definitions for actionable metrics to hold providers accountable to. Performed statistical and regression analyses in Python & Excel for individual Medical Directors to make data driven decisions to drive the goal of 50% reduction in opioid usage. Co-authored a paper to be publishing by JAMA (expected Fall 2018) outlining our progress as a Task Force and the statistical methods we used to measure our progress.Authored an ad hoc statistical analysis of partner Skilled Nursing Facilities and their average cost for commonly occurring procedures, which was used in meetings with representatives from the SNF’s and provided statistical support to C-Suite executives for presentations around reimbursement negotiations with major insurance carriers.Create and maintain SSIS packages that are deployed for our ETL processes. Packages include everything from simple data transfers to custom developed C# and Python Script tasks.

Empirically define, model (activity networks) and document (Visio and Word) business processes and work flows.Create & manage monthly / quarterly / ad-hoc reporting scorecards for North American, EMEA & APAC region heads of sales and Business Management executives including visuals and summary points to ‘tell the story’ of what is going on.Use MS PowerPivot/Query/BI in conjunction with MS Excel/Word/Power Point to create cloud based Business Intelligence solutions.Perform ad-hoc statistical analyses on various cohorts pertaining to asset, flows, opportunities and meetings data.Partner with members of our London office to develop, construct and implement automated processes to increase efficiency of a resource constrained team. Worked with team members to help build their technical skills.Mine large sets of internal and external data for valuable insights for the Region Heads, Relationship Managers and other analysts to use in daily/strategic business operations as well as presentations to T. Rowe Price executives and current clients / prospects. Developing a prototype MongoDB NoSQL database to store and analyze large volumes of unstructured data. Prototype is developed with C#. The application gathers unstructured files such as Excel, CSV and other text files, transforms the data to BSON and loads the data to the MongoDB database.Serving as SME for GIS Enterprise DataMart (DataMart is built in SQL Server 2008).

Furthered the USIS Broker Dealer strategy by deriving segments characterizing current Financial Advisor clients based on how well they exhibit ‘Home Office’ subscription behavior. Furthered the analysis by applying the models to create a data driven service model for Sales Associates to follow. Integrated the results and techniques into TIBCO Spotfire utilizing the TERR engine.Worked closely with Head Data Architect in creating data models to store Investment Performance Data migrated from the MIDR database into the ENT_Report DataMart. Partnered with lead ETL developers to ensure the proper business logic was followed when transferring the data.Authored various forms of documentation including but not limited to requirements for business intent, analysis visualization and system architecture, business process mapping, KPI metric decompositions and guided analysis of the business challenge at hand.Developed a prototype designed to demonstrate the predictive analytics capabilities Spotfire with R and how they can be applied to the business units we partnered with. Techniques currently included are: ANOVA, Linear Regression, Logistic Regression, Classification / Clustering Analysis, Principal Component & Factor Analysis.

Tutored MBA & Ph. D. students in graduate level Decision Management, Operations Research and Linear Programming courses. Provided Mathematics and Statistical consultation for graduate students' research papers / dissertations. Created VBA code to perform Monte Carlo simulations of incentive/disincentive rates for use in Management Engineering research. Used Excel to create an easy to use simulation analysis tool for the Management Engineering doctoral student I worked with. The book would run the simulations, analyze the results and create document ready narratives, charts and tables.

Co-created ERP budgeting system for Naval Reserve Client.Gathered, organized and documented requirements for ERP budgeting systems.Created Excel Add-In applications to perform monthly queries from the MAJIC ERP budgeting system.Created step by step user manuals for clients for the budgeting system we built.Created database structural flow chart for the ERP system designed for our Naval Reserve client.

Vetted Management Appraised Valuations of Illiquid Unmarketable Securities.Coded VBA applications to gather, classify and analyze equity price data, calculate historical volatilities and intrinsically value options using the Black-Scholes arbitrage free pricing method.Used Bloomberg & Reuters data feeds to gather spot and forward rates to construct yield curves for credit default swap valuation as well as calculating implied default rates of investment grade bonds.Selected to a 5 member team to design a revamped analysis system to be employed globally by all FTS pricing sites. Created MS Power Point training modules and programs for department wide use. Subjects included Excel, Visual Basic and TIS (Templeton Information System). Trainings contained portions dedicated to how the three programs can be used together symbiotically to increase productivity. New hire programs included notes & examinations for users along with notes for manager’s points of emphasis to newly hired individuals.Organized and managed quarterly employee reviews. Created performance metrics and review templates. These metrics were then used to quantify year-end performance bonuses and raises.Consistently received outperforming ratings (Year-end performance rating was 4.0 out of 5).

Classified and analyzed new security additions to the pricing database, specifically complex securities, to ensure proper documentation and valuation within the context of the Pricing Departments conventions.Designed policy and procedure enhancements to address operational inefficiencies and risks, daily NAV calculation precision as well as data integrity for securities held in multiple Franklin Templeton fund families.Member of Fair Value Team. • Prepared Market Level Fair Value success metrics and memorandums circulated to the Fair Valuation board members. I also developed a VB application to perform the memo creation process, decreasing completion time by two-thirds while reducing manual entry risk.Coded multiple macros to enhance the daily workflow processes. • Intrinsic Bond Valuation – Collected daily yields of debt instruments over an assigned benchmark to derive a risk free rate plus basis point spread intrinsic value of over 1500 individual bonds on a daily basis to monitor for mispriced debt instruments. • Activity Completion Time Simulator – Used observed daily completion times of tasks performed in the 4 – 6 pm “hot spot” of the mutual fund pricing day to simulate the long-run average.Gathered, classified, formatted & analyzed security price data to develop “smarter” operation exception thresholds.Used Bloomberg to efficiently gather data to model valuations for illiquid securities via financial models. Utilized a variety of statistical methods e.g. OLS & conditional probability to estimate unknown parameters for valuation models.Chosen twice (3rd overall) as a top contributor of all Operations employees to participate in roundtable discussions with members of Executive Management. Received year-end performance ratings of 4.0 out of 5; demonstrating that I consistently outperformed what was expected in all core and peripheral duties.

Reviewed, analyzed and resolved pricing exceptions on a daily basis for equity, fixed income and derivative securities. Area of specialty was foreign corporate & sovereign debt as well as domestic distressed debt and structured products.Utilized Microsoft Excel and Visual Basic by creating macros to minimize manual input risk while increasing productivity of the Fort Lauderdale pricing site. Selected as a top contributor of all Operations employees to participate in roundtable discussions with members of executive management.

Contract employee for Franklin Templeton Services in their Security Maintenance department.