Why on earth would they host orchestrating applications in Russia? AWS, Hetzner, DigitalOcesn work just fine.
In Russia it's higher latency, way more expensive and there are lots of connectivity issues. Its not like Russia have ministry of social network botting that require to host it within country.
Reason why these bot networks are effective and hard to find is because it's not some "KGB" minister respobsibility.
Botnets, malware, botting in Russia is someone small criminal enterprises with a lot of people doing it independently. They might be affilated to government or not at all - might be they just sell services to highest bidder.
The business case reason Russia would have a central orchestrator in a Russian data center is that they want centralized control. They are unlikely to delegate control of the information sphere to another, or allow a huge random mass of uncoordinated individuals just “post positive stuff about Russia”.
Putin seems to very much like control, and I doubt he would delegate control over foreign propaganda to agents outside Russia.
This is especially true if you want to spread a lie. Lies have to be coordinate, the truth does not.
Latency is irrelevant for social media posting, Western providers might identify and suspend such an operation overnight, and no, hosting in Russia is actually a lot cheaper.
Well, I had context and a nice technical explanation for it ready, but my longer comments get instaflagged since the account is new, so I'll let this cool off for a bit first.
In short: it's a site to track the movies and shows you've watched and quickly figure out what to watch next rather than scrolling Netflix squares for two hours.
Bit of a backstory: I've been running GamesGraph, a game backlog/discovery site, solo since 2019. The core problem there (you've got this huge pile of stuff from Steam sales, what do you even play next? Instant decision paralysis) translates pretty well to films and shows, so I decided to fork the whole thing. Same engine and same interface, except it's pointed at films and shows. I rely on the TMDB API and I've seeded the ML model with ~32 million real ratings from MovieLens, so it's based on collaborative filtering rather than genre tags.
Everything you rate tells the site how to rank your watchlist, so the backlog sorts itself and gently nudges you to either watch whatever is at the top or consciously push it to the bottom. The discovery queue has a one-at-a-time format with plenty of useful filters for power users.
Be aware: the TV side is admittedly a bit raw, because I've only had a few testers so far and this needs more data for accurate recommendations, so you might want to start from movies. It's completely free, no account needed unless you want to save your progress, and I'm not planning to monetize. Poke around and tell me where it gets your taste wrong or what's confusing.
They don't decide what to post independently. An orchestrating application is hosted somewhere, and that somewhere is Russia.