In January, we launched our 2025 machine learning training with a new series of webinars called Discover Anemoi. The series introduces the features and usage of the Anemoi AI weather forecasting framework.
Anemoi is a collaborative, open-source framework designed for developing machine learning-based weather forecasting models. Launched in 2024 as a joint effort between European meteorological services, its goal is to provide the essential building blocks for training state-of-the-art data-driven forecasting models and running them in an operational setting. The framework comprises a suite of Python-based tools for working with datasets, defining model architectures, training models, and running inference, among other capabilities. ECMWF’s AI-driven forecasting model, the AIFS, is trained using Anemoi.
The Discover Anemoi webinar series aims to help meteorologists and scientists across Europe and beyond stay up to date with the rapid developments in Anemoi.
Our Introduction to Anemoi session provided an overview of the ecosystem and its core components, while follow-up webinars focused on datasets, graph-based approaches, training, and inference. Each session explored specific features of the framework with hands-on examples and demos, showing how Anemoi can be leveraged for operational weather forecasting and research.
Webinars allow us to reach a broad audience with short, focused training sessions, and the response and participation was highly encouraging. More than 500 people from over 60 countries registered for this series, highlighting the growing role of machine learning in advancing weather forecasting capabilities worldwide.
Replay the webinar series
If you missed the live sessions, don’t worry; all webinar recordings, slides, and example materials are available on the event page.
We’re always looking to improve our training. If you have suggestions or feedback, we'd love to hear from you. You can contact the training team by email.