Brad Ferguson, PhD

Steered a cross-functional team of data scientists, business intelligence analysts, insights analysts, and product analysts to deliver data-driven solutions, achieving operational excellence and strategic goals. Facilitated the integration of data science into business decisions and operational processes, leading to the seamless assimilation of insights into decision-making. Developed and implemented predictive models and machine learning algorithms to forecast business outcomes and steer strategic planning. Advocated for the use of Decision Intelligence across the organization, ensuring data and insights are applied appropriately to inform business decisions. Managed the design and implementation of BI tools and dashboards, providing actionable insights to stakeholders and improving business performance.

I directed the application of machine learning and statistical methods to address Digicert's key concerns. I constructed highly precise classification models, empowering the sales and marketing teams to effectively identify customers at heightened risk of service migration.

During my tenure at Domo, my principal obligation encompassed the integration of data science functionalities into the Domo product suite. In collaboration with the software engineering and development teams, I architected applications and tools that enabled non-data scientist users to execute machine learning and statistical procedures on their data. I authored the mathematical code underpinning all these methods and meticulously supervised their comprehensive deployment.

At IQVIA, I offered expert statistical consultation on clinical trials focusing on oncology, rheumatoid arthritis, and diabetes. My core duties included conducting thorough literature reviews and meta-analyses, facilitating the determination of accurate sample sizes and the development of optimal experimental designs. Moreover, I consistently executed Monte Carlo simulations to scrutinize various experimental hypotheses.

Assisted in the translation of enrollment methodology from R to C#. Derived additional enrollment methodology such as site level enrollment caps, delayed enrollment start dates, and heterogeneous enrollment rates for both baseline and interim projections. Provided internal and external consulting for enrollment planning in various indications. Developed phase 1 dose-escalation forecasting tool to assist in phase 1 planning. Developed cost models for high level cost predictions.

Teaching assistant for Stat 516 - Graduate Level Statistics for Engineers

Worked 20 hours a week under a lead statistician in drug development areas including nonlinear mixed effects modeling, Bayesian hierarchical modeling, and clinical trial simulations.

Calibrated atmospheric computer models using Bayesian gaussian process models. Ran LFM-MIX magnetosphere model on NCAR super computer and developed code to post-process results.

Worked 20 hours aweek developing Bayesian methods for use of Gaussian processes to binary data. Implemented Bayesian adaptive experimental design for cost efficient testing.

Dissertation on "Auxiliary Bootstrap Methods" which allows the application of the bootstrap in the world of Big Data.

Thesis on "Threat Detector Testing Using Bayesian Gaussian Process Models"