Ecology-informed AI for advancing biodiversity science

About Me

Originally trained in computer science, I dispersed into the niche of ecological science through a PhD in biodiversity modeling, and I’ve since settled as a computational ecologist. My research lies at the intersection of ecology, statistics, and artificial intelligence, where I design models to explain biodiversity patterns, uncover ecological processes, and forecast ecosystem responses under global change.

Research

My work bridges ecological theory and machine learning to develop frugal, scalable, and trustworthy models that can handle high-dimensional, imbalanced biodiversity data. I combine ecological surveys, functional and habitat data, and multi-modal environmental information (climate, remote sensing, citizen science) to advance both theory and applied conservation.

My previous work includes:

• Predictive AI for community-level modeling: Developing deep joint species and community-level distribution models.
• Trustworthy and frugal modeling: Explainable AI for multi-species distribution models, foundation models for frugal learning and imbalance aware training.
• Ecological structure learning: Inferring ecological association from large-scale biodiversity data using dependency networks and species representation learning.
• Leveraging ecological structure in biodiversity models: Integrating ecological network and macroecological properties within predictive models.
• Remote sensing of ecosystems: Leveraging high-resolution Earth Observation data and foundation models to map habitats, monitor biodiversity, and support large-scale modeling.
• Applied conservation support: Contributing to European and national projects to deliver modeling tools that guide biodiversity assessments and conservation planning.

I collaborate across ecology, computer science, and conservation practice, supervising students from diverse backgrounds and co-leading projects with academic and non-academic partners.

Hobbies

Outside of research, I explore the natural world by foot 🥾, fin 🤿, or bike 🚴 wherever curiosity leads - ideally not all at once. I occasionally play 🏀 basketball, just to remind myself that not all nets are ecological or neural.

Contact

Feel free to reach out if you’re interested in collaboration, data sharing, or simply want to talk about AI for good.

📧 [firstname].[lastname]@univ-grenoble-alpes.fr