Data Theoretic is an independent digital lab with experience helping clients research, develop, and deploy cutting-edge AI projects and building AI teams in industry.

About the Team

Patrick Harrison serves as an AI consultant, researcher, and educator. Before starting Data Theoretic, Patrick spent eight years working in machine learning at a major financial intelligence company. As the first data scientist in the organization, he partnered with senior leadership to identify high-impact opportunities to deploy AI within the business. After several pilot projects delivered measurable business value, Patrick had the opportunity to start and lead the company’s AI Engineering team. He led a recruiting effort that resulted in hiring sixteen high-powered data scientists and machine learning engineers. He managed the design and implementation of the organization’s infrastructure, tools, and workflows to support machine learning research, development, and production operations. Along the way, he spent time working at many stages of the engineering career ladder: from a solo data scientist, to a tech lead with a small team, to an engineering manager with a larger team, to a department head responsible for multiple AI development teams. He has practical experience working on business applications using many areas of machine learning and natural language processing, including text classification, information extraction, document recognition, search, recommendations, probabilistic data matching, customer segmentation, time-series forecasting, churn prediction, and more. Patrick’s group became an incubator for technology talent: many of the team members he recruited, hired, and led have since gone on to hold senior technical positions at companies like Meta (Facebook), Microsoft, Capital One, hedge funds, and a variety of smaller, tech-forward organizations.

Concurrently with his role in industry, Patrick was invited to teach at the University of Virginia Darden School of Business. He served as an adjunct faculty member and one of the primary instructors for the Data Science in Business course, part of the Master of Business Administration program. Data Science in Business provided MBA students with a survey of data science and machine learning techniques and taught students how to implement their ideas in code with Python.

Patrick is very active in the broader AI and technology communities. He was invited to join the Data Science Leaders Advisory Council, a network of top data science leaders and decision-makers convened by Domino Data Lab to share knowledge about the trends, issues, and practices influencing the success of data science teams in their companies and institutions. He has given numerous conference talks, including appearances at high-profile industry conferences like the Strata Data Conference, PyData, and the Rev Data Science Leaders Summit. He served as one of the founding organizers for the Applied Machine Learning Conference in Charlottesville, Virginia. The 2019 Applied Machine Learning Conference attracted 500+ attendees and 59 speakers from industry, academia, and the non-profit sector. On the local level, he serves as one of the core organizers for the  PyData Pittsburgh meetup and previously organized the Charlottesville Data Science Group.

Patrick received a master’s degree in systems engineering and a bachelor’s degree in economics from the University of Virginia. His graduate research focused on applying agent-based modeling and simulation to better understand distributed decision-making and social, economic, and ecological outcomes in complex socio-environmental systems.

Patrick is currently based in Pittsburgh. If he’s not at his desk, you’ll probably find him running on one of the many excellent trails in Frick Park.