From satellite images
to decisions on the ground.
I’m an AI Lead at Sunesis in Ljubljana, working on solar forecasting, distributed learning, and intelligent infrastructure. My projects explore how machine learning can support the critical operational decisions in energy and computing systems.
Solar Energy, 2025 · Solar forecasting
A transformer approach to solar forecasting that combines satellite images, weather data, and historical production.
Kubernetes · Intelligent autoscaling
A Kubernetes-native system for intelligent autoscaling based on custom service metrics. The second version accompanies a SoftwareX submission, currently in review.
Brdo pri Kranju, Slovenia · 2025
I presented my master’s thesis work on forecasting solar power production using satellite images at the ENFIELD workshop on Green AI and sustainable energy systems, co-organized by Iskraemeco.
SoftwareX, in review code
Slovenia · Solar forecasting
Solar forecasting system for day-ahead trading for the Slovenian transmission system operator.
Master’s degree in Data Science
Bachelor’s degree in Electrical Engineering
For the Reproducibility Challenge 2021, I reproduced the empirical results of an ICML paper on coherent probabilistic forecasting for hierarchical time series.
Third place with my team in the 2021 DataScience@UL-FRI competition, working on click prediction with a problem proposed by Zemanta.
Silver medal at the 2017 Macedonian Mathematics Olympiad.