Jev acts as final decision layer in a crypto trading engine

Brendan Playford @BrendanPlayfordOriginally posted 2026.09.17

Open original on X

I built a trading engine with Jev and a 12B trained model to predict outcomes and get an edge trading crypto markets Here’s how it works: A trained probabilistic model generates thousands of possible price paths across multiple timeframes We combine those paths with live candles, ATR, EMA, RSI, volume, and market structure We generate valid long and short candidates with: → Entry zone → Stop loss → TP1 + TP2 → Risk/reward → Expiry → Invalidation rules Jev reviews the typed evidence and chooses: LONG, SHORT, or NO TRADE Reply to get free API/MCP access