When we watch a school of fish from above, we do not see thousands of individual trajectories or a complex network of interactions. We see a cohesive body whose boundary continuously ripples and deforms. These dynamics raise a simple question: could the fluctuating shapes of animal groups reveal the laws governing their collective motion? This project aims to turn the shape fluctuations of wild fish schools into a quantitative probe of collective dynamics. Combining aerial drone observations with statistical data analysis, we will investigate how boundary deformations emerge, propagate and decay. Recordings of different species in the Caribbean, the Atlantic and Indian oceans will allow us to test whether these fluctuations follow common physical laws across contrasting environments. To explain these observations, we will combine simulations of interacting fish with hydrodynamic theory. In particular, we will investigate whether milling schools, in which fish circulate around a common centre, can be described as droplets of active fluids. We will then test whether the parameters governing collective motion can be inferred from boundary fluctuations alone, without tracking every fish. Your internship will focus on a specific part of this program, with the possibility of continuing into a PhD. Depending on your interests and skills, your work can combine data analysis, numerical simulations and theoretical modelling, with opportunities to contribute to field observations.