ML / AI Researcher & Data Scientist — Basketball Player and Lineup Modeling

We are looking for a strong machine learning researcher, AI researcher, data scientist, or statistical modeler to help with an applied NBA modeling project. The focus is on building better models around player impact, lineup performance, substitutions, coaching rotations, offensive/defensive metrics, and live game context. We want to better understand how the players currently on the court change team strength, and how those signals can improve broader game prediction models. This is not a dashboard or reporting role. We are looking for someone strong in statistics, probability, machine learning, Bayesian modeling, feature engineering, model validation, and modern AI workflows who can apply those skills to a complex real-world sports modeling problem. Strong candidates should have experience with some combination of: Python Machine learning Data science / statistical modeling Bayesian modeling Predictive modeling Time-series or sequential modeling Feature engineering Model validation and backtesting Probability and uncertainty estimation Regression, classification, or hierarchical models Neural networks or deep learning LLMs, AI agents, or AI-assisted research workflows Automated data analysis or model experimentation Working with messy real-world datasets Sports analytics experience is helpful, especially with basketball, but the most important requirement is strong technical modeling ability. Relevant basketball experience may include: NBA play-by-play or possession-level data Player impact modeling Lineup analysis On/off metrics Adjusted plus-minus, RAPM, EPM, BPM, or similar concepts Rotation modeling Win probability or game prediction models Offensive and defensive efficiency modeling Possible work may include: Designing player and lineup-level features Building statistical, machine learning, or AI-assisted model prototypes Analyzing how lineups change team strength Modeling substitution and rotation patterns Evaluating which player-level metrics add predictive value Creating Python notebooks and research outputs Using modern AI tools or agents to accelerate research, data processing, and model iteration Recommending how to integrate player and lineup signals into a broader prediction model We are open to starting with a smaller paid research project or trial assignment, then expanding the scope if there is a strong fit. Please apply with relevant machine learning, AI research, data science, statistics, Bayesian modeling, or sports analytics work. Include links to GitHub, Kaggle, papers, notebooks, dashboards, models, or prior projects if available. Also include a short note on how you would approach modeling player and lineup impact in NBA games.