Optimizing agent setups for desired outcomes

[simonausten]

I'm keen to learn if anyone has encountered and tackled the challenge of optimizing agent setups for specific outcomes.

My objective is to devise a setup that delivers the most effective results for a certain scenario. I'm interested in understanding if anyone has worked on a similar problem, and if so, what strategies were employed? Topics such as genetic programming, reinforcement learning, or other optimization methods might be of particular interest.