Poverty killer by CEO MWAS

Project Outline: AI-Powered Trading Bot for Deriv  I am looking to develop an AI-based trading bot for Deriv, specifically focused on continuous indices with a primary strategy centered on digit prediction. Below are the specific requirements and conditions for the bot’s operation:  1. Objective and Strategy  Market: Deriv's continuous indices, focusing on digit prediction.  Strategy: The bot should predict that the last digit of the tick will be below a certain number, with adjustments based on trading results.  Primary Condition: Start with trades predicting under 9 by default.  Loss-Triggered Adjustment:  If the bot experiences a loss on an "under 9" trade, the next trade should predict under 5.  If the bot loses again after switching to "under 5," it can either stay on "under 5" or revert back to "under 9," depending on which approach is optimized to reduce consecutive losses.   Reversion Strategy: After a winning trade, the bot should reset and trade under 9 as the primary condition.   2. AI Model Requirements  Model Type: A supervised learning model, like LSTM, to analyze past price patterns and predict whether the last digit will be under 9 or under 5. If feasible, a reinforcement learning aspect would be ideal, allowing the bot to refine its strategy based on trading performance.  Training Data: Use historical and real-time data, focusing on price, volume, and any relevant technical indicators. Feature engineering should include moving averages, volatility, and RSI to enhance predictive accuracy.   3. Data Requirements  Data Sources: The bot should use both historical and real-time data on price and volume from Deriv’s continuous indices.  Data Frequency: Updates should occur every minute to ensure the bot has recent data for predictions.  Feature Engineering: Include relevant technical indicators, such as moving averages and volatility indicators, to improve the model’s predictive power.   4. Risk Management Rules  Position Sizing: Limit each trade’s position size to 2% of the total account balance.  Stop-Loss and Take-Profit: Set a stop-loss after a maximum of 6 consecutive losses to prevent major drawdowns. Implement a take-profit target as well to lock in gains after reaching a predetermined profit threshold.  Maximum Drawdown: Cap the maximum allowable drawdown at 10% to protect capital.   5. Backtesting and Validation  Backtesting: The model should be thoroughly backtested across historical data and stress-tested under various market conditions to ensure robustness.  Performance Metrics: Evaluate the bot using metrics such as win rate, Sharpe ratio, maximum drawdown, and average trade duration. Target a win rate above 60% and a Sharpe ratio of at least 1.5.   6. Deployment and Execution  Platform: The bot should operate on the Deriv platform, connecting via API for real-time trade execution.  Execution Speed: Ensure minimal latency to avoid slippage and allow for efficient trade entries and exits.  Monitoring and Alerts: Set up notifications to alert me if the bot reaches either its profit target or maximum drawdown, as well as alerts for key events like consecutive losses.  Custom Notification: On each notification (such as at startup, major trading events, or milestones like take-profit or stop-loss), the bot should display the message: “Poverty Killer Bot V1 by CEO MWAS ”.   7. Maintenance and Updates  Retraining: The bot should retrain on recent data monthly to adapt to changing market conditions and maintain performance.  Continuous Monitoring: Implement a monitoring system to check performance regularly and allow for quick adjustments if it underperforms.