Recent developments in orchestration and machine learning have made network automation more feasible, allowing the transition from prone-to-error, time consuming, manual manipulations to fast and refined automated responses in areas such as network security and management. This article investigates the capabilities of an RL agent to learn how to automatically distribute prefixes, correct undesired network behaviours and increase network resiliency and security. Our work focuses on network saturation, approaching the problem of network responsiveness in massive content delivery scenarios. Additionally, we propose a platform architecture to continuously monitor and deploy actions to the network.