Denis S Willett

Research Seminar

Data-Driven: Applying AI to Sustainable Systems Science

Artificial Intelligence holds tremendous potential for engendering step changes in our ability to manage agricultural and natural systems. Realizing this potential is challenging due to the interdisciplinary nature of the core ingredients necessary to support production AI applications at scale. Here, we will dive into applied AI in the context of data-driven science. First we will discuss the key ingredients necessary to support production AI while using case studies to highlight successful implementation of AI solutions when these ingredients are present. Next, we will discuss what the future holds with a focus on untapped potential for AI applications in agricultural and managed natural systems. A key theme emphasized throughout the presentation will be the interdisciplinary nature of systems science for sustainable agriculture.

Presentation Access

This Research Seminar is designed to be both accessible and reproducible. You can preview the content and follow along during the presentation here. In addition, all of the code used to generate the presentation and content is publicly available on GitLab here.

This presentation was developed using RevealJS and Quarto

Teaching Seminar

Ethics of Applied AI in Data-Driven Science

Artificial intelligence and data science are extremely powerful tools, the full consequences of which are only beginning to be appreciated by both practitioners and society at large. Here, we will take a case-study driven approach with real-world examples to highlight ethical pitfalls of applying AI tools at scale. We will focus both on broader impacts of personal decision making and the personal ramifications of such decisions for data scientists. Using examples from Industry and Government, students will step into the shoes of data scientists in ethically precarious situations then learn about the actual decisions data-scientists made in those same situations. Following these case studies, we will take a step back and discuss how these lessons learned map to work in sustainable agricultural systems.

Background

This teaching seminar is an excerpt from the Earth Systems Data Science in the Cloud intensive offered at the North Carolina Institute for Climate Studies. This activity is offered as the introduction to Module 2: Data Product Development | Statistics/AI & ML after students have gone through Exploratory Data Analysis and Data Preprocessing. More information about the course can be found on the course homepage.

Presentation Access

This Teaching Seminar is designed to be both accessible and reproducible. You can preview the content and follow along during the presentation here. In addition, all of the code used to generate the presentation and content is publicly available on GitLab here.

This presentation was developed using RevealJS and Quarto

Chalk Talk

Principles, Practices, Proposals, Platforms and People: Building Sustainable Systems AI at NCSU

This Chalk Talk will focus on the key ingredients needed to support interdisciplinary applied AI research in sustainable agricultural systems at NC State University. A strong emphasis will be placed on the role of diversity, equity, and inclusion in creating collaborative high-performing teams that build community. We will also discuss strategies for building organizational capacity, the fundamental interplay between research and teaching, how this work fits into the broader NCSU ecosystem, and avenues for funding.

Denis S Willett

Dr. Willett joined NCICS in 2021 as a Research Scholar, Software Engineer contributing to the NOAA Open Data Dissemination Program. His work focuses on leveraging cloud technologies for the development of data processing and machine learning pipelines. Prior to NCICS, he developed analytics and machine learning pipelines for agriculture as faculty at Cornell AgriTech and as a data scientist across roles in industry, academia, and the federal government. Denis did his PhD in Entomology and Nematology at University of Florida and his undergraduate and masters work in Earth Systems at Stanford University.

Denis specializes in designing and deploying full-stack, cloud-native data science solutions to intractable technical problems across domains. At NCICS, Denis works on petabyte scale multi-cloud data transfer pipelines, creating pipelines to furnish analytics-ready environmental data, and building production machine learning platforms applying MLOps principles to automate model development and deployment. Previous work has involved building high-throughput metabolomics and volatilomics phenotyping platforms, developing agricultural sensing and monitoring systems for pests and pathogens, expanding the capabilities of scientific instrumentation using machine learning and optimization, building prediction platforms that materials science engineers could use to leverage machine learning for better product design, scaling IOT ingest and analytics pipelines, implementing workforce development plans for data science teams, and strategic positioning planning for bioinformatics companies.

A key component of this and previous work has been technical advocacy in organizations adopting machine learning and cloud technologies. In working with and advising start-ups, Fortune 500 companies, Government Agencies, and Universities, architecting change from the C-Suite to the technical teams has been a priority with efforts focused on education, opportunity, and capacity building. A key part of this is building and leading diverse teams to implement change at the strategic and technical level.

Elements

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';