Machine Learning Consultant & Python Developer for Crypto Trading Signal Analysis (Monthly Engagement)

I am looking for an experienced machine learning consultant and Python programmer to help me analyze 3 years of BTC/USDT 5-minute candlestick data. This is a *personal project* focused on evaluating and improving trading signal efficiency.

What I'm trying to achieve:

1. Supervised Learning: Analyze the effectiveness of my existing entry and exit signals
2. Unsupervised Learning: Discover new potential signals from the data

Current indicators I have:
- One custom proprietary indicator (with 4 sub-components)
- RSI
- Multiple EMAs

The goal is to build a "constellation" of indicators/parameters that show high probability of success. You'll help me develop the feature set and identify which combinations work best.

Technical skills required:
- Knowledge of technical analysis, indicators etc
- Proficiency with XGBoost, LightGBM for gradient boosting models
- Optuna or Bayesian optimization for hyperparameter tuning
- SHAP for model interpretability and feature importance
- Monte Carlo simulations for robustness testing
- VectorBT for backtesting and performance analysis
- PyCaret for rapid ML prototyping and model comparison
- MLFinLab for financial machine learning techniques (purging, embargo, combinatorial purged CV)
- SkopeRules for interpretable rule extraction
- Genetic programming for feature engineering and strategy evolution
- Familiarity with StrategyQuant X concepts is a plus

What I need from you:
- Guidance on ML strategies and techniques suitable for this dataset
- Help with hyperparameter tuning and feature selection
- Python code that I can run locally on my machine for validation
- Clear communication and willingness to brainstorm ideas together
- Patience to explain concepts and discuss approaches
- Structured and logical thinking : ability to break down complex problems methodically
- Comfortable working with large datasets without feeling overwhelmed. This will involve significant data processing and feature engineering at scale

Critical: Risk management & pitfalls
You should advise on and help implement safeguards against:
- Overfitting (using proper cross-validation, purging, embargo)
- Look-ahead bias
- Data leakage
- Multiple comparison bias
- Curve fitting traps
- Out-of-sample validation protocols

Engagement details:
- This is an ongoing monthly engagement, not a one-time project
- You'll write clean, documented Python code and hand it over to me to run/validate on the dataset on my local machine
- After we establish the parameters, we'll likely move to building a paper trading bot (with Telegram integration)

Ideal candidate:
- Strong written communication skills
- Open to discussion and collaborative brainstorming
- Knowledge of trading/financial markets is a big plus but not mandatory
- Experience with time-series data and ML is essential

Note: I posted a similar requirement earlier but couldn't finalize due to work commitments. This is a personal project I'm serious about completing, so I'm reposting with the intent to close quickly this time.

Please share your relevant experience and monthly rate. I'm happy to discuss the technical details further if you have questions.

Looking forward to working with someone who enjoys solving real problems with data.