> So just read and judge every piece of text entered into the whole platform No. Not at all. We're software engineers, right? Make the machines do the heavy

> So just read and judge every piece of text entered into the whole platform No. Not at all. We're software engineers, right? Make the machines do the heavy lifting. Create a simple keyword based alerting mechanism for known inflammatory language in the languages that you support. There are resources to help you make this easier in multiple languages, i.e. https://hatespeechdata.com/ https://hatespeechdata.com/ Add a moderation system for your users to flag abusive content. Until you grow to a huge size, or unless your platform is courting violence promoting content, there should be a relatively small amount of content you need to manually review and remove.