Forex AI System Enhancement
Budget: $25 – $50 USD
Project Status Summary: Forex AI Trading System
Current Situation:
Our Forex trading system is in the advanced optimization stage, employing AI-driven logic and multiple technical indicators to generate intraday and short-term trade signals.
Indicators in Use:
Signals Overlay (SO)
Oscillator Matrix (OSC)
Price Action Concepts (PAC)
Divergence signals from multiple indicators
Poki (being replaced with Lorentzian Classifier)
Trading Timeframes:
3 Minutes (3M)
5 Minutes (5M)
15 Minutes (15M)
1 Hour (1H)
Performance:
Signals Generated: 24–30 per week
Current Accuracy: 82%
Target Accuracy: 90%+
The system is designed to deliver frequent, high-precision signals for consistent weekly trading opportunities.
Key Workflow:
Signal Processing Flow:
All technical indicator outputs (SO, OSC, PAC, Divergence, LC) are sent in real-time to GPT, which analyzes these combined signals and applies decision logic to determine whether a valid trade signal should be generated.
Once validated, GPT sends approved trade signals onward for execution or user action.
Key Issues & Objectives:
Replace Poki with Lorentzian Classifier (LC):
Upgrade indicator stack to improve signal quality and achieve higher accuracy.
Improve Trend & Divergence Logic:
Current model often fails to act on divergence signals in strong trends, leading to false signals or missed reversals.
Refine GPT Prompt Logic:
Revise AI prompt instructions to enhance decision-making, particularly when divergence suggests a possible trend reversal.
Accuracy Calculation Methodology:
Accuracy is calculated based on whether the price reaches a predefined profit target of $4 (equivalent to 40 pips) from the entry point.
Signals are evaluated only after they achieve or fail to achieve this default target.
Next Steps:
Integrate Lorentzian Classifier and fully replace Poki.
Conduct thorough backtesting to validate performance improvements.
Refine AI prompt logic to ensure correct prioritization of divergence in trending markets.
Maintain weekly output of 24–30 signals with a strict 90%+ accuracy target.
---
Summary Statement:
The system is stable and producing a consistent volume of 24–30 signals per week, with current accuracy at 82%. The next focus is on integrating the Lorentzian Classifier and refining AI prompt logic to improve divergence handling in trending conditions across all active timeframes (3M, 5M, 15M, 1H).
All indicator outputs are analyzed by GPT before issuing any trade signal, ensuring that decision logic is centralized and consistent.
Accuracy tracking is standardized: signals must successfully achieve a $4 profit from entry to qualify as accurate.
Current Situation:
Our Forex trading system is in the advanced optimization stage, employing AI-driven logic and multiple technical indicators to generate intraday and short-term trade signals.
Indicators in Use:
Signals Overlay (SO)
Oscillator Matrix (OSC)
Price Action Concepts (PAC)
Divergence signals from multiple indicators
Poki (being replaced with Lorentzian Classifier)
Trading Timeframes:
3 Minutes (3M)
5 Minutes (5M)
15 Minutes (15M)
1 Hour (1H)
Performance:
Signals Generated: 24–30 per week
Current Accuracy: 82%
Target Accuracy: 90%+
The system is designed to deliver frequent, high-precision signals for consistent weekly trading opportunities.
Key Workflow:
Signal Processing Flow:
All technical indicator outputs (SO, OSC, PAC, Divergence, LC) are sent in real-time to GPT, which analyzes these combined signals and applies decision logic to determine whether a valid trade signal should be generated.
Once validated, GPT sends approved trade signals onward for execution or user action.
Key Issues & Objectives:
Replace Poki with Lorentzian Classifier (LC):
Upgrade indicator stack to improve signal quality and achieve higher accuracy.
Improve Trend & Divergence Logic:
Current model often fails to act on divergence signals in strong trends, leading to false signals or missed reversals.
Refine GPT Prompt Logic:
Revise AI prompt instructions to enhance decision-making, particularly when divergence suggests a possible trend reversal.
Accuracy Calculation Methodology:
Accuracy is calculated based on whether the price reaches a predefined profit target of $4 (equivalent to 40 pips) from the entry point.
Signals are evaluated only after they achieve or fail to achieve this default target.
Next Steps:
Integrate Lorentzian Classifier and fully replace Poki.
Conduct thorough backtesting to validate performance improvements.
Refine AI prompt logic to ensure correct prioritization of divergence in trending markets.
Maintain weekly output of 24–30 signals with a strict 90%+ accuracy target.
---
Summary Statement:
The system is stable and producing a consistent volume of 24–30 signals per week, with current accuracy at 82%. The next focus is on integrating the Lorentzian Classifier and refining AI prompt logic to improve divergence handling in trending conditions across all active timeframes (3M, 5M, 15M, 1H).
All indicator outputs are analyzed by GPT before issuing any trade signal, ensuring that decision logic is centralized and consistent.
Accuracy tracking is standardized: signals must successfully achieve a $4 profit from entry to qualify as accurate.