As a consequence of the increasing instability and uncertainty around migration patterns, and sometimes populist fears about border control in Europe, governments increasingly turn to new information infrastructures, sources of digital data and forms of analysis like predictive algorithms. We ask: what are the implications of using algorithms as redistribution keys for migrants? Drawing on the interdisciplinary convergences between Migration Studies and Science and Technology Studies, the aim of this project is to follow the implementation of one particular algorithm, developed as a pilot project by the Migration Agency in order to increase statistical precision of predicting migration to Sweden. Using a multi sited ethnography, we also trace its connections with databases available in institutions across Germany to understand how algorithms are produced, implemented and received by different national and transnational institutions and actors (developers, case-managers, migrants, policy-makers). This proposal seeks to reveal coordinations mediated by algorithms in different locations to understand how integration policies depend on information infrastructures and institutional processes.