Super ethical computer challenge
A multi agent simulator written in Python implemented as a distributed application. It employs MPI for inter process communication and can handle a large number of independent agents by using a multiple threaded non-preemptive design.
Agents are implemented using the classic "Intelligent Agent" design, i.e. a layered behavior approach that has perception and actuators.
A simple reactive agent implementation was created to showcase the simulation.
Unbeknownst to me at the time, this network architecture is used a lot in large multiplayer games.
There are other smaller classical AI projects in my <Github>.