An Ethereum whale tracker that learned which wallets to watch
Most on-chain data tools are slow, expensive, or retrospective — by the time a retail trader sees that a whale moved $40M into a token, the move is already priced in. I wanted a system that watched Ethereum in real time, identified meaningful wallet activity, scored every event against what the market actually did afterward, and learned which wallets were worth following.
A Go service running on its own infrastructure that ingests Ethereum blocks in real time, profiles 177,000+ wallets across nearly 2M whale events, scores each move 0–100 based on signal quality, labels known entities (CEXs, bridges, funds), and surfaces the top performers to a Telegram channel. The system self-grades — every prediction gets compared to the price action that followed, so the pool of 'smart money' wallets (2,500 and counting) emerges from data, not from my opinion.
$6T in on-chain flow observed. 2,500 wallets worth watching, identified automatically.