This is a smart product. The mechanics work, and the UX is mostly good (a bit sluggish, and a few quirks I'd love to see improved*). Overall I think it is worthwhile. The AI comments are generally good, but the real issue lies with which posts you're actively commenting on. Identifying by keyword is too one-dimentional. I'd love to see more AI logic built in, specifically in terms of checking if the post you're commenting on is (a) from your country, and (b) truly relevant to your brand. Without that, you're potentially commenting on irrelevant or potentially very brand-damaging posts based on a keyword and a closed set of banned words (helpful, but very very far from fool-proof). That said, based on that, I don't think I'd ever trust this to go on autopilot, which is a bummer. Also, I wish there was an opportunity to provide the AI more context about your brand to help it better draft the comments (and ideally, to achieve the point mentioned above). As the cost of AI API tokens continues to fall, maybe we'll get a bump in function in the future. Otherwise, this is currently a good, but not great platform.

*Some additional improvements:
-Ability to edit the keyword vs delete it and start over (see point below)
-Global banned words, in addition to keyword specific banned words (or at least don't make me rebuild the list if I change the keyword)
-Bulk delete comments (or purge "deleted" or denied posts)
-Auto-blacklist your already blocked social users (prob not possible with X.com API)