Middle Data Scientist building and maintaining time-series and tabular predictive models end to end: EDA, feature engineering, designing and tuning neural networks in PyTorch (with the Darts library as a plus), evaluating and improving production models, and writing production-grade Python, including internal ML libraries.