Why Field Observation Still Matters in the Age of Foundation Models
On the gap between what AI can classify and what humans can learn by looking closely.
Building AI for ecological observation, interpretation, and human connection.
I work at the intersection of artificial intelligence, ecology, and field observation — exploring how technology can deepen the way humans understand plants, landscapes, and living systems.
Selected Work
Projects
AI Naturalist Companion
Rewild
An AI system that guides people to observe nature more carefully — not just identifying species, but helping users understand why a plant is identified through visible features like leaf shape, bark texture, and habitat context.
Computer VisionXAIEcologyNext.jsLLM
Remote-Sensing Phenology
NOVA Invasives
Detecting kudzu, phragmites, and other invasive species in Northern Virginia by exploiting their phenological signatures in Sentinel-2 time series — processed entirely on Google Earth Engine.
GEESentinel-2NDVIPhenologyPython
Rainfall & Water Quality Analysis
Rewild Water
Time-series analysis exploring the relationship between precipitation events and water quality indicators in DC-area watersheds — turbidity, dissolved oxygen, pH, and discharge patterns.
Time SeriesHydrologyPythonData Science
Coming soon
More projects in progress
Exploring ecological visualization, seasonal data patterns, and AI-assisted field observation tools.
Essays & Field Notes
Writing
Coming soon
Why Explainability Matters in Plant Identification
Moving beyond 'what is this plant' to 'how can I learn to see the differences myself.'
What Suburban Ecology Taught Me About AI Observation
Lessons from watching osprey, invasive species, and seasonal change in places most people overlook.
Ecological Attention
Observation Archive
Seasonal landscapes, wetlands, birds, and suburban ecology — documenting the living systems around us.
Photo 1
Photo 2
Photo 3
Photo 4
Photo 5
Photo 6
Photo 7
Photo 8
Photos coming soon — follow @ivy.observes for field observations.