Coluna is a framework to implement decomposition approaches to tackle block-structured mixed-integer programs (MIP). The platform offers a library of functionalities to develop a so-called Branch-and-Price-and-Cut algorithm and to handle extended formulation for combinatorial optimisation problems. Coluna is developed on an open-source basis and hence can engage innovations and constant updates; it is made to serve as a solid socle to researchers building further algorithmic progress. This open source project is carried by Atoptima, a spin-off from Inria/CNRS/University of Bordeaux 1, in collaboration with academic partners. Another important feature of Coluna.jl is that it is completely implemented in Julia, a recent dynamic language that provides high productivity and high performance. Choosing Julia to develop Coluna.jl was motivated by striking the best trade-off between efficiency and ease to share co-developments under a simple syntax to facilitate updates and extension of the framework.