Все заказы
Правила
Техническое задание
Technical Project Description: AI-Enhanced Adaptive Trading Expert Advisor
I want to commission the development of a high-performance AI-powered Expert Advisor (EA) designed for MetaTrader 5, optimized for indices (e.g., Volatility 75 Index), gold (XAU/USD), and crypto pairs (BTC/USD). The EA should intelligently adapt to market volatility using machine learning or AI-based decision-making to refine its entry and exit logic over time based on trade outcomes and market conditions.
The EA must combine price action and indicator-based strategies, including EMA crossover, RSI confirmation, MACD trend filtering, and Bollinger Band volatility detection. It should also include engulfing and pin bar candle detection for precision entries. All trade management functions—Stop Loss, Take Profit, Breakeven, and Trailing Stop—should be dynamically calculated using the ATR (Average True Range) indicator to suit different volatility conditions.
Risk management should be handled using a percentage-based lot sizing system (1–2% of equity per trade), with features like daily trade limits, equity protection, and spread/volatility filters to avoid unfavorable market conditions. The EA must also support multi-timeframe confirmation, checking higher timeframes for trend alignment before opening trades.
To enhance decision-making, the EA should integrate a lightweight AI model (such as a reinforcement-learning module or TensorFlow Lite model) that continuously evaluates trade success patterns, dynamically adjusting the aggressiveness, entry precision, and indicator thresholds over time.
An on-screen dashboard should display all key statistics (balance, equity, open trades, profit, spread, trade status, AI confidence score, and trend direction). The system should include data logging for all trade actions, market states, and AI adjustments for continuous backtesting and improvement.
Additionally, the EA should have client-specific license control—each version of the EA should include an expiration date or user ID binding so that it can only function for authorized clients or within a specified duration. This can be managed via local key verification or a simple online validation script.
The final EA should be fully optimized, fast-executing, and compatible with both manual backtesting and live trading environments. It should run efficiently with minimal CPU/memory usage and support auto-updating configuration parameters based on market regime detection (trending, ranging, high volatility, etc.).