.png&w=256&q=75)
🤓 Latest Submissions
The Future of Autonomous Bot Trading: JuventusAI
JuventusAI is an autonomous algorithmic trading framework designed to solve financial AI hallucination through a multi-stage validation pipeline and strict risk engineering. Built around an Alpaca API execution engine, the platform combines quantitative technical indicators (EMA, RSI, ATR), Google's Gemma 4 31B LLM for trade validation, and a hardcoded Deterministic Risk Gate to ensure zero-hallucination execution. Before executing live, the strategy undergoes rigorous backtesting and parameter optimization using Python's Optuna library. By running hyperparameter optimization trials across historical market data, Optuna fine-tunes technical moving average windows, confluence buy/exit thresholds, and volatility multipliers for peak risk-adjusted performance. Once deployed, JuventusAI continuously monitors price action. When quantitative confluence triggers a setup, the Gemma 4 31B model evaluates market context to filter out false breakouts. Approved setups pass to the Risk Gate, which mathematically calculates position sizing (capping risk at 1.5% portfolio equity) and sets ATR-based stops. Orders are submitted to Alpaca via automated Bracket Orders with dual-layer exit mechanics—combining server-side Stop-Loss/Take-Profit orders with dynamic, signal-driven market exits—providing disciplined, 24/7 autonomous trading.
4 Sep 2026
