Reputation-weighted predictions on independent music
Est. 20268 attested2 open calls
A friend sends you a SoundCloud link; the track blows up a year later; nobody remembers they called it. Scenius makes that track record portable and verifiable — then turns it into a signal that funds and agents pay to read, over x402 on Base.
A human picks an emerging SoundCloud track and predicts whether it clears a play threshold by a deadline. Making the call costs a small USDC stake, paid over x402 — skin in the game, and the spam tax that keeps the board honest.
Play counts are snapshotted at call time and rechecked by a weekly cron. Right or wrong is settled against live SoundCloud numbers — not opinion, not an opaque model.
Each resolved call updates an exponential moving average of Brier score. Good calls compound; luck decays. Every outcome is attested onchain via EAS on Base.
Weight every open call by its caller's reputation and you get a credibility signal per artist — the read of the people who have actually been right before.
That signal is a metered API. Autonomous agents pay per query in USDC over x402 to pull the watchlist. Humans supply the taste; machines pay to consume it — and never make a call themselves.

suave (feat. Jane Remover)
AKRIILA
0xd929...2ba0 predicts will hit 10,000 streams in 1 week

suave (feat. Jane Remover)
AKRIILA
0xd929...2ba0 predicts will hit 10,000 streams in 1 week