Sktime

Sktime is an open source Python framework for time series machine learning. It puts forecasting, classification, anomaly detection, feature extraction and probabilistic prediction behind one consistent API, so a team can fit and compare models without learning a new interface for each method. The library covers more than 500 models, spanning classical statistics, machine learning, deep learning and modern AI foundation models. Swapping one algorithm for another is a small edit, and users can bring their own estimators, transformers or adapters from a closed codebase and run them through the same interface. The site points to use cases such as retail demand planning, hourly energy load forecasting, predictive maintenance on sensor data and volatility modeling in finance. The project is aimed at data scientists, researchers and engineering teams who work with temporal data. It is released under the permissive BSD 3 Clause license, governed by its community and free to use. A separate company offers enterprise services around it, including implementation and production integration, support with service levels, forward deployed engineers, and training and consulting. The library itself carries no price. Enterprise engagements are arranged through a contact form, and the vendor does not publish prices for them.