Trading systems researched in public. Historical real-money losses stay visible. New capital remains behind future-only, executable-price and drawdown gates; a heartbeat never gets mislabeled as a working strategy.
A disciplined, high-conviction bot — only its best reads, a few a day, holding winners and scaling out with discipline. Every position shown live.
Two bots race on $200 paper each, live on Hyperliquid — an ML directional + delta-neutral funding-carry bot vs a bot that copies the best on-chain traders. Same stake, head-to-head, net of fees & funding.
Quoted both sides of Polymarket's 5-minute BTC and ETH up-down markets. Captured the spread, hedged the fill — and the math never closed.
The retired real-SOL KOL-copy experiment. Its complete loss remains public while a new launch model is tested on uncontaminated, future-only data.
Hunts dormant coins that wake back up — buys confirmed breakouts from quiet bases, real SOL. Wins and losses shown straight.
Our machine-learning model trading a real $50 fully autonomously — the public legit check. Every trade settles on-chain.
The real-money copy experiment. Chain-verified fills, redemptions and open marks are shown below; it stays paused while its replacement is evaluated after costs and delay.
The second real-SOL meme experiment. Its current book is negative, so execution is halted while the six-month launch lake is rebuilt into leakage-free walk-forward tests.
The launch research lake. It contains hundreds of thousands of labeled launches, but only seven currently qualify as catchable $1M runners — too few for an honest live claim.