Quantitative ML Researcher Needed for Predictive Backtesting Project

Description I’m looking for a quantitative data scientist / machine learning researcher to test whether a historical structured dataset contains a statistically significant predictive edge. The first phase is not to build software, a dashboard, or an AI chatbot. The goal is to run a rigorous research/backtesting exercise and determine whether the data contains useful signal. Deliverables 1. Clean and prepare the dataset 2. Define the target variable clearly 3. Build a simple baseline model 4. Build one or two predictive ML models 5. Run a proper out-of-sample backtest 6. Identify feature importance / drivers of prediction 7. Explain whether the model has genuine predictive edge or not 8. Provide a clean Jupyter Notebook or Python project with clear documentation Required skills Python, pandas, NumPy, scikit-learn, XGBoost / LightGBM / CatBoost, statistics, feature engineering, backtesting, model validation, SQL. Preferred background Quantitative finance, sports betting, credit risk, insurance, fraud detection, pricing models, forecasting, or other structured-data predictive modelling. Important I am not looking for a generic AI chatbot, LLM agent, automation specialist, or dashboard developer. This is a predictive modelling and backtesting project. The ideal candidate should be able to think critically, avoid overfitting/data leakage, and honestly assess whether the model works or does not work. Initial project scope This is a fixed-price proof-of-concept. If the first phase shows promising results, there may be follow-on work to expand the dataset, improve the model, and eventually build a usable research platform.