Amrinder Jakhar
Agronomist and applied data scientist. I build and test predictive models from crop, soil and sensor data, from the plot to the satellite, so farmers can make better nitrogen and water decisions.

From field plots to models
I am a Ph.D. candidate in Crop and Soil Sciences at the University of Georgia, advised by Dr. Leonardo M. Bastos, and I defended my dissertation in October 2026. My work compares proximal, UAV and satellite sensing for in-season nitrogen recommendations in irrigated corn, and I led a global meta-analysis of sensor-based variable-rate nitrogen in maize.
Before UGA, my M.S. at Mississippi State University focused on soil health and water quality: how biochar and gypsum change soil properties and nutrient losses in Mississippi Delta cotton and soybean.
I fit machine learning models (gradient boosting, random forests, explainable AI), work mainly in R with Python, and share my analysis code openly. My goal is to build AI tools for soil and crop decisions and test them against field data.
Recent
- Presented an oral talk and a poster at CANVAS 2026 in Portland, OR.
- Successfully defended my Ph.D. dissertation at the University of Georgia.
- 2nd place, NxS Hackathon, and 3rd place poster at the Nitrogen Use Efficiency Workshop.
- Invited speaker, Brazil-Global Digital Agriculture Workshop, UNESP Jaboticabal.
- Meta-analysis on sensor-based variable-rate nitrogen published in Precision Agriculture.
- 1st place, Bayer Graduate Hackathon Challenge (Team DataGenix), CANVAS 2025.
What I work on
Sensor-based nitrogen management
In-season, variable-rate nitrogen for corn: when to apply, how much, and which sensor to trust.
Remote sensing across platforms
Comparing proximal canopy sensors, UAV multispectral and satellite imagery over the same plots.
Machine learning & explainable AI
Predictive models from crop, soil, weather and sensor data that are tested against field outcomes.
Soil health & water quality
My M.S. work at Mississippi State: how soil amendments change soil physics and nutrient losses to water.