Dynamic application migration between cloud providers

A system comprising:

• a memory coupled to a processor;
• a migration service that executes on the processor, uses the memory, and is configured for:
• receiving budget information of a containerized application deployed with a set of containers to a set of machine instances to a first cloud provider service of a set of cloud provider services;
• receiving pricing information from each cloud provider service of the set of cloud provider services, wherein the set of cloud provider services includes the first cloud provider service and a second cloud provider service;
• receiving performance information of the containerized application from the first cloud provider service;
• generating an output vector from a machine learning model, wherein the machine learning model uses the pricing information and the performance information to generate the output vector;
• and
• migrating the containerized application from the first cloud provider service to the second cloud provider service based on the output vector by removing a first container from the first cloud provider service and deploying a second container to the second cloud provider service, wherein the first container and the second container each perform a same function.

A non-transitory computer readable medium comprising computer readable program code for:

• receiving budget information of a containerized application deployed with a set of containers to a first cloud provider service of a set of cloud provider services;
• receiving pricing information from each cloud provider service of the set of cloud provider services, wherein the set of cloud provider services includes the first cloud provider service and a second cloud provider service;
• receiving performance information of the containerized application from the first cloud provider service;
• generating an output vector from a machine learning model, wherein the machine learning model uses the pricing information and the performance information to generate the output vector;
• determining a first cloud provider service cost and a second cloud provider service cost based on the output vector and the pricing information;
• and
• migrating the containerized application from the first cloud provider service to the second cloud provider service based on the output vector, the first cloud provider service cost, and the second cloud provider service cost by removing a first container from the first cloud provider service and deploying a second container to the second cloud provider service, wherein the first container and the second container each perform a same function.