You're right -- we scrape a good deal of startups, but sometimes we get duplicates, and although we try to de-duplicate via a lot of different methods, sometimes they sneak through.

This one should be fixed now, but there's probably some other examples people will find as well, and we catch and fix them when we see them.