Tâche terminée
Temps d'exécution 5 jours
Commentaires du client
Great work! A highly professional developer.
Commentaires de l'employé
Very considerate and give you the best offer.
Spécifications
Upgrade Existing MT5 CatBoost Angle EA → Add Self-Learning Engine + Multi-Symbol + Prop-Firm Risk Layer (with 5-Year High-Quality Tick Backtests)
Project Overview
I already have a fully functional MT5 EA based on a CatBoost Angle Strategy (3-model ensemble).
This project is NOT about rewriting from scratch.
I need an experienced Senior MQL5 Developer to upgrade and extend the EA’s architecture by integrating:
A Self-Training Engine (pattern-score reinforcement learning)
Multi-Symbol independent learning (6 symbols)
A full Prop-Firm Challenge Risk-Control Layer
High-quality (99% modeling) tick-data backtests for 5+ years
Clean, modular OOP architecture
CSV-based model persistence per symbol
All CatBoost logic, angle features, and ensemble prediction must be kept exactly as they are.
Your job is to extend the EA with new modules and integrate them cleanly.
Scope of Work
1. Integrate the Self-Training Engine (Reinforcement-Style, MT5 Internal)
No Python needed.
All learning is inside MQL5.
Includes:
Feature builder
Feature binning → pattern hashing
Pattern table per symbol
Online score update (win/loss smoothing + decay)
MAE/MFE tracking
Model persistence (CSV)
Learning must run in parallel with CatBoost, not replacing it.
Final Decision = CatBoost Prediction + Pattern Score + Risk Layer
2. Multi-Symbol Learning Engine
Must support the following symbols independently:
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Each symbol has its own:
Pattern table
MAE/MFE
Model CSV
Score history
Learning cycle
No cross-contamination allowed.
3. Prop-Firm Challenge Risk Layer
Must include:
Account-Level
Daily loss limit
Max drawdown limit
Equity guard
Account shutdown mode
Execution Level
Spread filter
Slippage filter
News/time blocked windows
Min seconds between trades (anti-HFT)
Max lots per symbol
Max total exposure
Strategy-Level
Swing mode (minimum holding time)
Risk Layer must wrap the trading engine:
CatBoost → Self-Training → RiskEngine.AllowNewTrade → Execution
4. High-Quality Tick-Data Backtests (Mandatory)
Modeling Quality: ≥ 99%
Symbols (≥ 5 years each):
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Timeframes:
M15
H1
Files required per backtest:
HTML report
Equity curve PNG
Monthly returns table
Raw trade history CSV
Tick-data source info
Modeling quality screenshot
5. Architecture Requirements
Must follow a clean module design:
[ Prop-Firm Risk Layer ]
↓ allow/deny
[ CatBoost Predictions ]
↓
[ Self-Training Engine ]
↓ Feature Builder
↓ Pattern Hash
↓ Pattern Table (per symbol)
↓ MAE/MFE Tracking
↓ CSV Persistence
No monolithic OnTick code.
Clean classes required.
Deliverables
Updated EA (.mq5 + .ex5)
Self-Training Engine (integrated with existing CatBoost logic)
Multi-Symbol learning engine
Prop-firm-safe risk layer
Model persistence per symbol
All 12 required backtests (6 symbols × 2 TF × 5 years)
Documentation
Clean, extensible OOP architecture
Project Overview
I already have a fully functional MT5 EA based on a CatBoost Angle Strategy (3-model ensemble).
This project is NOT about rewriting from scratch.
I need an experienced Senior MQL5 Developer to upgrade and extend the EA’s architecture by integrating:
A Self-Training Engine (pattern-score reinforcement learning)
Multi-Symbol independent learning (6 symbols)
A full Prop-Firm Challenge Risk-Control Layer
High-quality (99% modeling) tick-data backtests for 5+ years
Clean, modular OOP architecture
CSV-based model persistence per symbol
All CatBoost logic, angle features, and ensemble prediction must be kept exactly as they are.
Your job is to extend the EA with new modules and integrate them cleanly.
Scope of Work
1. Integrate the Self-Training Engine (Reinforcement-Style, MT5 Internal)
No Python needed.
All learning is inside MQL5.
Includes:
Feature builder
Feature binning → pattern hashing
Pattern table per symbol
Online score update (win/loss smoothing + decay)
MAE/MFE tracking
Model persistence (CSV)
Learning must run in parallel with CatBoost, not replacing it.
Final Decision = CatBoost Prediction + Pattern Score + Risk Layer
2. Multi-Symbol Learning Engine
Must support the following symbols independently:
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Each symbol has its own:
Pattern table
MAE/MFE
Model CSV
Score history
Learning cycle
No cross-contamination allowed.
3. Prop-Firm Challenge Risk Layer
Must include:
Account-Level
Daily loss limit
Max drawdown limit
Equity guard
Account shutdown mode
Execution Level
Spread filter
Slippage filter
News/time blocked windows
Min seconds between trades (anti-HFT)
Max lots per symbol
Max total exposure
Strategy-Level
Swing mode (minimum holding time)
Risk Layer must wrap the trading engine:
CatBoost → Self-Training → RiskEngine.AllowNewTrade → Execution
4. High-Quality Tick-Data Backtests (Mandatory)
Modeling Quality: ≥ 99%
Symbols (≥ 5 years each):
XAUUSD
EURUSD
GBPUSD
USDJPY
US30
DAX40
Timeframes:
M15
H1
Files required per backtest:
HTML report
Equity curve PNG
Monthly returns table
Raw trade history CSV
Tick-data source info
Modeling quality screenshot
5. Architecture Requirements
Must follow a clean module design:
[ Prop-Firm Risk Layer ]
↓ allow/deny
[ CatBoost Predictions ]
↓
[ Self-Training Engine ]
↓ Feature Builder
↓ Pattern Hash
↓ Pattern Table (per symbol)
↓ MAE/MFE Tracking
↓ CSV Persistence
No monolithic OnTick code.
Clean classes required.
Deliverables
Updated EA (.mq5 + .ex5)
Self-Training Engine (integrated with existing CatBoost logic)
Multi-Symbol learning engine
Prop-firm-safe risk layer
Model persistence per symbol
All 12 required backtests (6 symbols × 2 TF × 5 years)
Documentation
Clean, extensible OOP architecture
Répondu
1
Évaluation
Projets
27
7%
Arbitrage
9
33%
/
33%
En retard
1
4%
Travail
2
Évaluation
Projets
8
0%
Arbitrage
3
33%
/
67%
En retard
4
50%
Gratuit
3
Évaluation
Projets
1
100%
Arbitrage
3
0%
/
100%
En retard
0
Gratuit
4
Évaluation
Projets
180
46%
Arbitrage
3
33%
/
33%
En retard
1
1%
Travail
5
Évaluation
Projets
3
33%
Arbitrage
2
0%
/
100%
En retard
0
Gratuit
6
Évaluation
Projets
58
34%
Arbitrage
15
27%
/
60%
En retard
1
2%
Travail
7
Évaluation
Projets
0
0%
Arbitrage
0
En retard
0
Gratuit
8
Évaluation
Projets
0
0%
Arbitrage
1
0%
/
100%
En retard
0
Gratuit
9
Évaluation
Projets
33
42%
Arbitrage
0
En retard
3
9%
Gratuit
10
Évaluation
Projets
1
100%
Arbitrage
0
En retard
0
Gratuit
11
Évaluation
Projets
2
0%
Arbitrage
1
0%
/
100%
En retard
1
50%
Gratuit
Publié : 4 codes
12
Évaluation
Projets
33
27%
Arbitrage
20
10%
/
50%
En retard
11
33%
Gratuit
13
Évaluation
Projets
553
50%
Arbitrage
57
40%
/
37%
En retard
227
41%
Travail
14
Évaluation
Projets
0
0%
Arbitrage
0
En retard
0
Gratuit
15
Évaluation
Projets
2
0%
Arbitrage
4
25%
/
50%
En retard
1
50%
Gratuit
16
Évaluation
Projets
29
34%
Arbitrage
4
50%
/
25%
En retard
5
17%
Travail
17
Évaluation
Projets
3
0%
Arbitrage
1
100%
/
0%
En retard
0
Gratuit
18
Évaluation
Projets
13
38%
Arbitrage
1
0%
/
100%
En retard
1
8%
Gratuit
19
Évaluation
Projets
1
0%
Arbitrage
0
En retard
0
Gratuit
20
Évaluation
Projets
0
0%
Arbitrage
0
En retard
0
Gratuit
Publié : 3 codes
21
Évaluation
Projets
0
0%
Arbitrage
0
En retard
0
Gratuit
Commandes similaires
A good trend predicting indicator is the one which can identify the trend change as soon as it happens on the chart. when a new candle is formed it should tell whether its going to go up or down. I have already seen a lot of repainting trend predictors so if your indicator is repainting then please don't bother contacting. I would like to see the demo version and then if satisfied , I would want the source code too
Custom MT5 Price action multi-pattern EA
30 - 100 USD
Hey, I am looking for an experienced MQL5 developer to build a price-action EA for MT5 - XAUUSD and EURUSD on the 30M timeframe. Features required: Pattern recognition based on 3-candle structures and wick ratios. Proportional dynamic lot sizing based on account equity tiers. Time filters, Friday hard-close safeguard, and MT5 native calendar news filter (CPI/PPI lockout). FTMO risk controls (max spread, daily
Création d'un robot sur MQL5
630 - 640 USD
Bonjour, je recherche un développeur MQL5 expérimenté pour créer un Expert Advisor pour MetaTrader 5 basé sur une stratégie de trading intégrant des principes de gestion des risques rigoureux et d'intelligence financière. Le robot doit être capable de gérer plusieurs paires de devises et d'optimiser automatiquement les entrées et sorties en fonction de conditions de marché prédéfinies."
MT4/MT5 HFT EA us30
30 - 3000 USD
Hello everybody, I'm looking for an experienced MQL4/MQL5 developer to optimize a High-Frequency Trading (HFT) Expert Advisor for both MT4 and MT5. The EA performs consistently and profitably on demo accounts, but when it is run on Raw and Standard live accounts under what appear to be the same trading conditions, it begins generating losses. I do not have the original source code (.mq4/.mq5); I only have the
8 cap prop firm passing
30 - 3000 USD
I am looking for an experienced MQL4/MQL5 HFT developer to build or optimize a High-Frequency Trading (HFT) Expert Advisor that can successfully pass proprietary trading firm challenges and perform consistently under live trading conditions with brokers such as 8cap or BlackBull Markets . The developer should have proven experience with HFT execution, ultra-low-latency trading, broker execution, slippage, spreads
Hello, I am looking for an existing, fully backtestable MT5 Expert Advisor or EA portfolio designed specifically for passing an FTMO-style prop-firm evaluation. I am not looking for unrealistic profit claims or a visually perfect backtest. My priority is a statistically defensible probability of passing the challenge while respecting all risk limits. Primary objective The system should be designed for the current
MT4/MT5 HFT EA Live Trading
40 - 10000 USD
I have a High-Frequency Trading (HFT) Expert Advisor for both MT4 and MT5 designed primarily for US30 (Dow Jones Index) . The EA performs consistently and profitably on demo accounts, but when I run it on an IC Markets Raw or Standard live account, it starts generating losses under what appear to be the same trading conditions. At this time, I cannot provide the source code (.mq4/.mq5). I can only provide the
Version document : 1.0 Plateforme : TradingView Langage : Pine Script v6 Type : Indicateur d'analyse et d'aide à la décision (non-exécutant) 1. Présentation du projet Nom du produit ONYX SR V2 — Intelligent Support & Resistance Scalping System Objectif Créer un indicateur TradingView capable d'identifier automatiquement des opportunités de scalping basées sur : supports et résistances dynamiques ; action du prix ;
Fundamental News Trading Strategy
50 - 200 USD
I want to build a fundamental news trading bot that trade off economic news data, as we know every economic news data released always have effect on the asset associated with it, so this bot will take a trade instantly based on the news data released either to buy or sell, it will come with good money management and also SL and TP target based on price and pips value
TumiiFX
30 - 20000 USD
1. Use two EMAs: 20 and 50. If EMA 20 is above EMA 50 → uptrend (look for buys) If EMA 20 is below EMA 50 → downtrend (look for sells) 2. Wait for a pullback into the area between the two EMAs. - For buys: price must touch or move between EMA 20 and EMA 50 during the last few candles. - For stils: same idea, but in a downtrend. 3. Entry signal: Buy: a bullish engulfing candle in an uptrend after the pullback
Informations sur le projet
Budget
550 - 700 USD
Délais
à 10 jour(s)