Recomendations for Control AI

I am trying to figure out the best way to implement this idea. This question is similar and the answer is helpful, but not what i’m looking for:

I am trying to implement a second order AI to control my AI. I currently have the base level working what it is effectively a defend the castle type objective. I have two teams, one controlled by the player, and i’m trying to figure out how to control the enemy team. There can be a wide range of number of pawns controlled, but their overall movement is controlled with a Team AI, i am trying to implement the instructions for this team by priority, and only rally points to arbitrarily strategic positions.

As new features are added i want to be able to update the control AI rather than the team AI to prioritize either different locations or resources to be able to beat the player.

When this is done I would like to implement some type of machine learning to improve the base strategy. Are there any recommendations on how to implement the control? I am starting in Blueprints because i’m still becoming familiar with unreal but can implement as C++ after I finish figuring out how everything works.