Dungeons and Dragons has always been a game about telling stories and living out adventures, but some of the best stories are those that write themselves and have unexpected outcomes. To this end, I thought if the background characters were given a spark of autonomy it would enrich the playing experience.
For this I used a neural network, since it can be trained to act predictably, but can still be unpredictable in new situations. Neural networks are build from the ground, up, to use linear algebra, so I chose to use MatLab to program it. Matlab isn't free though, so you could use Octave or Python with a matrix algebra library, as well.
The software is on Github for anyone to download and enjoy......But not yet, since there's one more major bug to work out for updating the system.