The ODEON project continues to advance in its mission to simplify and scale the use of artificial intelligence in the energy sector, with new developments focused on improving the design and execution of AI workflows.
At the core of this progress is the evolution of the AI Designer, a key component of the ODEON ecosystem, which is now enabling the creation of reusable pipelines to automate common AI processes. This milestone marks a significant step towards making AI assets more accessible, interoperable, and easier to deploy across a diverse range of energy stakeholders—from technology providers to service integrators.
One of the main challenges addressed by this development is the complexity involved in moving from accessing contracted AI assets to effectively operationalising them. By standardising end-to-end workflows, the AI Designer reduces the technical and operational burden on users, allowing them to move more efficiently from acquisition to execution. The introduction of configurable pipeline templates minimises manual intervention while maintaining the flexibility required to support different providers, assets, and execution environments.
Currently, two main families of reusable pipelines are being developed. The first focuses on training workflows, covering the full process from model training and performance visualisation to the storage of trained models within the Federated Model Orchestration System (FMOS). The second addresses batch inference workflows, enabling users to run structured predictions using contracted models and input data.
In practical terms, this means that an energy forecasting model, for example, can clearly define its input requirements—such as historical time series data and forecasting parameters—and automatically generate outputs in a standardised format, such as time-indexed datasets with predicted values. This structured approach facilitates integration, reuse, and scalability across applications.
In parallel, the AI Designer has been integrated with ODEON’s user access and authentication framework, ensuring secure and consistent access to the platform. Together, these advancements aim to make AI workflows easier to configure and reuse within the ODEON ecosystem, avoiding isolated or fully custom implementations for each use case.