Abstract: Novel tools and techniques for machine learning based quality of experience optimization are provided. A system includes one or more network elements, an orchestrator, and a server. The server may further include a processor and non-transitory computer readable media comprising instructions executable by the processor to obtain telemetry information from a first protocol layer, obtain telemetry information from a second protocol layer, modify one or more attributes of the second protocol layer, observe a state of first protocol layer performance, assign a cost associated with changes to each of the one or more attributes of the second protocol layer, and optimize the first protocol layer performance based, at least in part, on the state of first protocol layer performance and the cost associated with the changes to one or more attributes of the second protocol layer. The orchestrator may be configured to modify the one or more attributes of the second protocol layer.