Developed containers and orchestrated their deployment in a secure environment using Docker Compose yaml and Dockerfiles.Created DAGS in Airflow for the batch processing of the XML documents.Developed templates using Jinja to batch process about 700k XSL-FO documentsUsed prompt engineering with Github Copilot to convert IBM Datastage jobs to Databricks and PySparkEngineered prompts to in Github Copilot to create unit tests based on code changes in git reposCreated ML models to predict mass closure incidents in a ServiceNow implementation using GCP BigQueryDeployed ML models to endpoints using Vertex AIDeveloped analysis of ITSM data to identify incident cascades initiated by other incidents within ServiceNowDeveloped scripts in python using Pandas to retrieve and deduplicate Nagios state change data at 5 minute intervals on a RHEL VMCreated models to predict the duplication of ServiceNow incidentsConducted analysis of ITSM data to optimize a ServiceNow implementationDeveloped a seminar on using Databricks and PySpark to analyze Energy Information Agency (EIA) electricity production dataBuilt data pipelines in Python to scrape EIA electricity production data and SEC log files into cloud storageImplemented and modified Terraform scripts and Jenkins to automate AWS deployments.Wrote and refactored SQL and bash shell scripts to transition a client from Cloudera to AWS EMR service using Hive.Developed a pilot model using a Ridge Classifier (L2 regularized logistic regression) and Random Forest from Scikit-Learn, and reporting in PowerBI to determine the root cause of patching failures in Windows 10 for a major client.Working on a remote AWS EC2 instance and Jupyter Lab, designed a metric for analyzing the technical viability (based on technical debt as measured by software end of life) and value (based on cost per server) of updating and migrating top 200 clients to cloud and other services.

Responsible for the validating, testing, and developing benchmarks for models used by the bank—including, but not limited to: DFAST, Trust Investment, and mortgage loss mitigation models.Used Excel to build optimization models that showed that following a simple portfolio allocation rule based on last month’s asset performance significantly improved the performance based on Sharpe Ratios of the Dynamic Asset Allocation fund.Developed benchmarking models to test the performance of Trust Investment’s equity selection models versus various alternatives. The analysis of the model validation group convinced the Trust Investments department to stop using a model developed by an outside vendor, saving the company $400k per year.Developed another suite of time-series models to predict loan charge-offs. The models showed that by changing the functional form used in estimating the models, reasonable forecasts of charge-off behavior could be obtained.Wrote R code that automated the analysis of model validation data using ggplot2 and dplyr.Developed a benchmarking model in Excel using non-linear optimization to determine the allocation of funds across asset classes based on diversification constraints. Under several scenarios the model outperformed the investments department’s Dynamic Asset Allocation Model to a significant degree using past data and measuring performance by the Sharpe Ratio.Managed and maintained model issues monitoring in MetricStream.Rewrote the reporting process, transitioning from a VBA based to a hybrid Excel and R based system. The original manual process used Excel spreadsheets and VBA code to process and compile reports that were distributed to model owners across the bank. The improved automated process imported the spreadsheets into R dataframes and generated Excel reports based on dplyr database join operations in R, significantly reducing the amount of hand processing to one or two operations across 20+ model owners.

Taught Money and Banking, Labor and Healthcare Economics, International Trade, Micro and Macro, and anything else in need of teaching. I also conducted applied econometric research in cigarette taxation, education, public choice, and electricity regulation; some university service as well.