🎯 Project Goal Micro-mobility systems like Citi Bike are highly sensitive to rapid shifts in demand. **Accurate short-term forecasts** help operators: - Rebalance bike docks efficiently - Plan proactive maintenance - Maximize bike availability for riders This project ingests historical 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 to enable live inference. An interactive **Streamlit dashboard** visualizes forecasts and station trends.