Hello, I'm
Geoinformatics Engineer & AI Researcher
Turning satellite data, deep learning, and a love for the night sky into meaningful insights about our planet — and beyond.
I'm a Master's student in Geoinformatics Engineering at Politecnico di Milano (expected Dec 2026), with a background in Environmental Engineering. My work sits at the intersection of Earth Observation, Deep Learning, and Big Data — using satellite imagery and neural networks to understand how our planet is changing.
I serve as Vice-President of the association managing the Ca' del Monte Astronomical Observatory, where I've also worked as a science communicator since 2018 — bringing the complexity of space down to Earth for general audiences.
Deep Learning framework for storm surge global prediction — exploring neural architectures that model ocean dynamics at planetary scale.
A full-stack toolkit for geospatial AI — from raw satellite bands to production-ready models.
End-to-end ML pipelines applied to real-world geospatial and scientific challenges.
Built an automated Google Earth Engine pipeline extracting 1,178 ten-band spectral signatures from 40 years of Landsat imagery to track glacier dynamics. Compared 1D-CNN, MLP, and Random Forest architectures on severely imbalanced geospatial data.
Designed a Bidirectional LSTM (2-layer, 156 units) for multi-class sequence classification using skeletal joints and clinical data. Applied rigorous user-based splitting and L1/L2 regularization to prevent data leakage.
Architected a robust vision ensemble combining EfficientNetV2, ConvNeXt, and ResNet152 for an 8-class image classification challenge. Tackled class imbalance through Focal Loss, Mixup/CutMix augmentation, and Test-Time Augmentation.
Fine-tuned the 300M-parameter Prithvi Vision Transformer (and Terramind) for a 10-class land cover segmentation task using only 100 training samples. Dynamically adapted RGB embeddings to process 7-band multispectral satellite chips.
Before I wrote my first line of Python, I was already staring at the sky. I serve as Vice-President of the association running the Ca' del Monte Astronomical Observatory, guiding visitors through the cosmos one eyepiece at a time.
Astrophotography is the practice of patience — long exposures, stacking frames, teasing signal from noise. Sound familiar? The same intuition I apply to satellite imagery processing started here, under dark skies, chasing photons from objects millions of light-years away.
It's also a constant reminder of why Earth Observation matters: seen from orbit, our planet is a fragile, luminous pale blue dot — worth understanding, worth protecting.
Always open to research collaborations, internship opportunities, or just a good conversation about space, AI, or our planet.