I am the Chief Technology Officer of Atlas AI, a small tech startup that generates geospatial intelligence on agricultural and economic trends across the developing world. If you want to lear more about our work and mission checkout the website: www.atlasai.co.
Before joining Atlas AI, I was a research scientist at Stanford University, where I worked with David Lobell at the Center on Food Security and the Environment (FSE), the Department of Earth System Science, and the Sustainability and Artificial Intelligence Lab. I worked on estimating crop yield, classifying crop type, and detecting management practices in several countries including USA, Zambia, Tanzania, Kenya, Uganda, and India. I still contribute, although marginally, to the group's research efforts.
I got my Ph.D. with Mike Goulden, in the Department of Earth System Science at UC Irvine. My dissertation was on tracking post-fire recovery of vegetation from satellite, but I devoted a lot of my time on illumination corrections algorithms for optical imagery. As a side research track, I explored the potential of gaming and photographic technology for field measurement with scientific standards.
A few selected papers, blog posts, and various news.
Using machine learning and satellite imagery to detect tillage practices in the North Central US.
Crop classification and yield estimates in Kenya and Tanzania using Sentinel imagery, SCYM, and machine learning.
Using machine learning and composite features from satellite imagery to classify land cover.
Using unsupervised techniques to predict crop type in the US.
Showing the scalability potential of our SCYM approach in three different countries.