1 comment

[ 4.4 ms ] story [ 11.7 ms ] thread
the bit about remembering past incidents is interesting. what happens when a new alert only looks like an old one? can you see which past incident it based the diagnosis on, and correct it if it matched wrong?