Distributed Constraint Optimization Problem (DCOP) is a useful framework for modeling many multi-agent coordination tasks. In the centralized case, several techniques have been developed to speed up constraint optimization solving. In particular, search can be improved by inference techniques, which identifies inconsistent values that can be removed from the problem. In this seminar we will see how to integrate these techniques in a distributed environment. In this process, agents only have partial information, so new operations must be performed in such a way that the problem remains globally coherent.