AutoTS is a Python library for rapidly deploying time series forecasts at scale. It provides dozens of forecasting models (naive, statistical, ML, and deep learning) with sklearn-style APIs that work directly on Pandas DataFrames. The library includes over 30 time series-specific transforms and automatically searches for optimal model configurations through genetic algorithms. It supports multivariate forecasting, probabilistic bounds, and horizontal/mosaic ensemble methods for scalable predictions across thousands of series.