🎯 Project Goal Traditional taxi services in urban areas like NYC are sensitive to fast-changing demand. **Accurate short-term forecasts** help operators: - Optimize fleet allocation and dispatching - Reduce passenger wait times - Improve operational efficiency This project ingests historical taxi trip data, aggregates it into hourly time series, engineers sliding-window features, and trains a LightGBM regressor to **predict number of rides in the next hour**. All features and trained models are tracked in [Hopsworks](https://www.hopsworks.ai/) for reproducibility and live inference. An interactive **Streamlit dashboard** visualizes forecasts and location trends.