How long have you been at EarthDaily Agro?
What’s your current position, and what does it entail?
I am a data scientist and as such I am responsible for creating analytics based on machine learning and deep learning. Specifically, I work on detecting objects in satellite images, such as irrigation systems (pivots). I also work on MLOps tools implementation to help model scaling across large areas of interest (AOIs) across the globe, such as countries and continents.
Have you grown professionally while on our team?
Working at EarthDaily Agro, I learned a lot about remote sensing, coming from an autonomous vehicles background and also about cloud. I also grew in my own skills by using MLOps as part of my day-to-day job.
What accomplishment during your time at EarthDaily Agro are you most proud of?
Pivot detection and field boundaries detection are my two main achievements, both in the model itself and, more importantly, in the scaling aspects, covering several countries/continents with both.
Why do you like your job? What do you find most interesting?
I like working in an organized team around data science subjects, and the challenge of applying machine learning/deep learning at a very large scale. Remote sensing is also a very interesting field, raising new challenges in image processing and model designs, notably in enriching the number of available features for model input with the multispectral bands.
What do you think of the EarthDaily Constellation arrival?
The daily revisit combined with a five-meter resolution will unlock whole new capabilities in detecting short-span events in a more efficient and reliable way. Many algorithms revolving around change detection or specific event detection will benefit from it. The five-meter resolution will improve the accuracy of all object detection algorithms, such as but not limited to irrigation system detection and field boundaries.
What are you passionate about outside of work?
Outside of work I enjoy painting miniatures and doing long walks with my dog!