Riot Toxicity detector?

Do they only use automatic system that look at profanity in game? I literally had someone who used racism/personal attack/suicide recommendation/physical threats/trying to get personal info such as where I live in post game screen.... Yet this person is still not banned. I was just curious if their system is dumb. If someone says fuck once or twice as an exclamation its still far less toxic than someone saying something along the lines of "your mother made a mistake giving a birth to you" ( for example) but first would probably be detectable by automatic system while second won't. Also, do you only review chat by personnel (if at all) if there is certain number of report criteria? cuz this seems abusable BW trolls.