I’m a founder of spacebayes, an early-stage Data-as-a-Service startup for crop yield forecasting and physical climate risk. We give trading desks, insurers, reinsurers and food companies early forecasts at national and county levels, together with uncertainty estimates. For that we’re building our own data infrastructure and fusing satellite (hyperspectral, optical and thermal), climate, weather and historical data.

I have six years of industry experience as a data scientist, where I focused on geospatial and Bayesian models, with applications in finance, supply chain and regional economic forecasting.

I have a PhD in Cognitive Science from Osnabrück University, where I worked on interpretable machine learning for real-world applications, like public health and environmental sciences. Previously, I got my bachelors in Physics at the University of Novi Sad, Serbia.

My work includes projects such as developing Bayesian hierarchical models to predict leaf area index (LAI) from reflectance spectra and modeling particulate matter (PM2.5) concentrations across counties in California.

I was always drawn to nature and my interest in science started with astronomy in our family’s back yard. During my bachelor studies, I’d every summer go to mountains with my colleagues from Petnica Science Center to observe meteor showers (talk about collecting data in real time!). I still like to watch meteor showers and other astronomical events, but I’m less focused on data, and more on the experience.

My hobby is analog photography, hiking and educating (aka bothering) my friends and family with David Lynch filmography.