Parallel requests

I would like the platform to support parallel requests to AI models. This change would significantly speed up processes that do not require sequential execution or have components that could benefit from parallel computation.