Clinical Assistant Professor of Digital Agriculture & Computational Biology
Department of Crop Sciences

ese @ illinois.edu

About Me

I am a Clinical Assistant Professor of Digital Agriculture & Computational Biology in the Department of Crop Sciences at the University of Illinois Urbana-Champaign. My research sits at the intersection of artificial intelligence, computational biology, quantitative genetics, sensing, azard and risk modeling, and complex systems.

My career has spanned academic research, biotechnology, agricultural R&D, entrepreneurship, and technology development. Before my current appointment, I was a Technical Project Lead at X, the Moonshot Factory, formerly Google[x], where I led interdisciplinary projects applying artificial intelligence to gene discovery and innovative uses of plants.

Earlier in my career, I led computational biology and quantitative genetics research in the agricultural biotechnology industry, including work at Benson Hill and Syngenta. I have contributed to the development and application of computational approaches for complex-trait dissection, genomic analysis, crop improvement, and predictive modeling.

At UIUC, I previously served as Associate Director of Educational Programs for the Center for Digital Agriculture and continue to develop research and educational programs connecting computation with agricultural and biological sciences. I am also the founder of Umbrella Genetics Foundation and Umbrella Genetics, LLC.

My Research Interest

My research focuses on understanding agricultural systems as complex, adaptive systems whose behavior emerges from interactions among genetics, environment, management, biological processes, technology, and human decision-making.

A central objective of my work is to develop computational approaches that integrate information across these scales rather than treating individual components of an agricultural system in isolation. I am particularly interested in how artificial intelligence, machine learning, simulation, sensing, and multimodal multiscale data integration can be used to characterize system state, predict outcomes, evaluate interventions, and support decision-making under uncertainty.

My group’s current research themes include:

• Integrative modeling of genotype × environment × management systems, including approaches for connecting biological mechanisms, environmental variation, management decisions, and crop performance.
• Artificial intelligence and machine learning for complex systems, including representation learning, reinforcement learning, multimodal modeling, and methods for discovering latent structure in biological, environmental, and spatial data.
• Generalized hazard and risk modeling with AI, developing computational frameworks that integrate hazard, exposure, vulnerability, spatial context, and uncertainty to characterize risk across diverse systems. This work explores how AI and multimodal data can support transferable approaches to hazard identification, risk profiling, forecasting, and intervention planning.
• Digital Fields and continuous multiscale environmental sensing, integrating remote sensing, proximal sensing, field experimentation, robotics, and 3D reconstruction in computational models to create dynamic representations of agricultural and environmental systems.
• Plant health and biosecurity, including multimodal sensing and AI-enabled approaches for early detection, characterization, risk assessment, and intervention.
• Climate resilience and sustainable systems, with an emphasis on understanding how biological, environmental, management, and hazard processes interact across spatial and temporal scales.
• Computational approaches to crop genetics and quantitative biology, building on my longstanding research in genetic architecture, association mapping, quantitative genetics, and predictive modeling for evaluating potential solutions for improvement of food systems.

Beyond these specific topics, I am braodly interested in looking for answers to the broader question: How can we construct computational representations of complex biological systems that are sufficiently predictive, interpretable, and actionable to help us design better interventions?

Research Philosophy

My work follows a compute-first philosophy: use computation, modeling, simulation, and artificial intelligence to explore hypotheses, integrate evidence, quantify uncertainty, evaluate alternatives, and identify promising interventions before committing resources to physical implementation.

Agricultural and environmental systems provide especially compelling settings for this approach because they operate across biological, physical, technological, economic, and human scales. Their complexity creates both a scientific challenge and an opportunity to develop methods for integrative modeling that can generalize across domains.

The long-term goal of my research is to develop computational frameworks that connect observations to models, models to predictions, predictions to risk, and predictions and risk to interventions—enabling complex systems to be studied not simply as collections of independent components, but as interacting systems whose behavior can be measured, learned, anticipated, and ultimately managed.

About Agriculture and Farming

Farmers are often unfairly accused of being resistant or being skeptical of new technologies, often out of a preference for more traditional methods or concerns about the potential negative effects of technological advancements.

In fact, this resistance is often due to gaps in technology dissemination and lack of subsidy structures and educational programs that would support and enable farming communities to afford, learn and adopt new practices.

Many farmers operate on slim margins, trying to balance the costs of production against unpredictable revenues. Farming margins being small is a result of a combination of factors, both inherent to agriculture and influenced by external market, economic, and policy forces. There are many factors that contribute to the current undesirable state of farming margins, which can best be considered as externalities, and can be modeled as a complex adaptive system.

Complexity theory, often associated with the broader field of complex systems, studies how relationships between parts give rise to the collective behaviors of a system and how the system interacts and forms relationships with its environment. This theory embraces the idea that complex phenomena emerge from a few simple behaviors and are governed by feedback loops, wherein the output is recycled as input. It acknowledges that complex systems are dynamic, nonlinear, adaptive, and sensitive to initial conditions, often exhibiting properties that cannot be predicted, reduced, or easily simplified.

By leveraging insights from complexity theory, stakeholders in agriculture can craft more effective strategies to introduce technological changes, ensuring that they are not only technically feasible but also socially and economically sustainable. This holistic approach acknowledges the intricate web of factors that contribute to technological inertia and provides a roadmap for navigating and overcoming these challenges.

My research program approaches agricultural systems as complex adaptive systems, using machine learning, artificial intelligence, simulation, and integrative modeling to understand the interacting forces that shape agricultural productivity, technology adoption, resilience, and farm-level outcomes. A central goal is to identify interventions that can reduce barriers to innovation while improving the economic and environmental sustainability of food production.

SanJose, CA 2023

Further reading on Complexity Theory and Complex Adaptive Systems

Complexity: A guided Tour by Melanie Mitchell

The Systems View of Life: A Unifying Vision by Capra & Luisi

Complexity Theory Takes Evolution to Another Level
by WIRED magazine

US Food and Agriculture System as a Complex Adaptive System

• /ɛlˈhɑːn/
  • /ɛ/ is pronounced like the “e” in “bed”.
  • /l/ is pronounced like the “l” in “love”.
  • /hɑːn/ has the “h” sound as in “hat”, followed by the “a” in “father” (stressed and elongated), and ends with the “n” sound as in “net”.
  • So, it would be pronounced like “el-HAHN”, with the emphasis on the second syllable.

• /ɛrˈsœz/
  • /ɛr/ represents the “Er” part, where “ɛ” is similar to the “e” in “bed” and “r” is the rolled or flapped r, common in many languages including Turkish.
  • /sœz/ represents the “söz” part, where “œ” is a sound that’s between the English “o” in “bird” and “a” in “bad.” The “z” is as in “zebra.”