Gaize is committed to safe workplaces and safe roads through the development of technology to provide evidence of impairment.

TeraDact provides information management technology to identify, tag and remove information, manually or automatically according to user-defined rules. TeraDact’s capabilities ensures the privacy and security of sensitive information including data residing in the cloud, unstructured data, freeform text inside documents and underlying metadata with an audit function that provides access control and logging.

* DataPhilly is the largest data-related Meetup in Philadelphia (over 5,000 members), covering topics related to predictive analytics, applied machine learning, big data, data warehousing and data science.* Assist in planning and organizing events.* Periodically give talks on Data Science. Previous topics include principal component analysis, neural networks, and the misuse of statistical techniques.* Help manage the group's web presence.

Managed Sidecar's data science team, providing hands-on mentorship and technical leadership. Responsible for facilitating the growth of the team's overall technical competence.Worked with others in leadership to improve workflow and efficiency across the organization. Worked with technical and non-technical leadership to meet business objectives. Ensured successful, timely completion of data science projects.Created, monitored, and drove adherence to OKRs.

Designed, built, tested and deployed machine learning systems for automating customer advertising campaigns in various online advertising channels. Primarily focused on automated bidding and automated keyword generation / evaluation.Built entire systems from the ground up: gathering data, training models, and finally building and deploying dockerized REST APIs to kubernetes clusters allowing for engineers to easily interface with machine learning models.

Sidecar is the only fully programmatic advertising technology that helps retailers get optimal results in Product Listing Ad channels like Google Shopping, Facebook Dynamic Product Ads, and Bing Shopping.

I tought an introductory artificial intelligence class and a class on agent-based systems several times. I also taught an introductory robotics course once. My thought on university level classes is that every course needs equal parts of the theoretical and the practical. I avoid any sort of memorization tests, preferring assignments containing written questions to demonstrate mastery of the concepts, and coding assignments to demonstrate the ability to put those concepts into practice.

My research focused on multi-agent systems and mobile ad-hoc networks. I authored or co-authered 32 publications. I originally went for an MS, but stayed for the PhD program. While there I spent some time consulting for the Naval Research Lab and being a visiting scholar in the Czech Republic. I also spent about a month riding around in the back of Humvees.

* Designed and implemented web scraper to detect abuse and other issues.* Designed and implemented system to import customer email address lists, utilizing several machine learning techniques to prevent abuse.* Designed and implemented machine learning algorithm to detect misleading emails.* Created classifier to predict customer churn.* Analyzed impact of forum spambots on our business.

* Worked with Czech scientists on a USARL project involving multi-agent planning.* Wrote software for multi-agent coordination using approximated distributed constraint reasoning to improve the scalability of their system.* Published 3 papers in collaboration with CTU colleagues.* Liaison between Drexel and CTU and continued collaboration after returning to the USA.

* Performed data analysis using Python on a project investigating coordination strategies in communications constrained environments.* Generated reports and data visualizations for the project manager, used to determine correlation between different strategies' features and outcomes.