Due to the success of internet and huge consuming services e.g. youtube, traffic passing through operators network is showing an exponential growth. Besides, the adopted flat rate charging policy reduces the Average Revenue Per User (ARPU) and makes the traditional actions such as the costly deployment of new equipments inefficient anymore. In order to overcome these challenges offload techniques at the RAN and CORE levels were proposed. However, network optimization via the management of the different offload functions could be long and complex. This is why, in this paper a self-optimization solution ensuring some level of human tasks automation is provided.