You can only know what is not original when you have a copy of what is original. Everyone kept telling us that this is an impossible mission. You have to first find a way of accessing that much data, store that data and go through that data, which grows like crazy every single day.
Yet, we did that. We download every day over 100 thousand new software projects and find very tiny code snippets at a scale you would say "impossible". This is the story of our startup, to know the software story of every software project made available to public.