{"id":32103,"date":"2026-07-16T16:30:52","date_gmt":"2026-07-16T14:30:52","guid":{"rendered":"https:\/\/www.ifabfoundation.org\/?post_type=project&p=32103"},"modified":"2026-07-28T11:00:36","modified_gmt":"2026-07-28T09:00:36","slug":"flaas-federated-learning-as-a-service","status":"publish","type":"project","link":"https:\/\/www.ifabfoundation.org\/en\/project\/flaas-federated-learning-as-a-service\/","title":{"rendered":"FLaaS: Federated Learning as a Service"},"content":{"rendered":"

National Centre for HPC, Big Data and Quantum Computing (ICSC), a project funded by the European Union \u2013 NextGenerationEU \u2013 and by Italy’s National Recovery and Resilience Plan (PNRR) \u2013 Mission 4, Component 2.<\/strong><\/p>\n

Objectives<\/h3>\n

FLaaS aims to offer a federated model training service, allowing users to submit training jobs through a simple interface, run on a distributed learning platform that includes a community of peers contributing their local data.<\/p>\n

This platform aims to demonstrate the technological feasibility and key advantages of a model in which organizations can collaborate on training machine learning models without directly sharing their own data. This will make it possible to build a variety of datasets without having to compromise the privacy of partners’ data.<\/p>\n

In addition, the system will be accessible and easy to use, so that users are able to set up and start model training efficiently and securely.<\/p>\n

Initial challenge<\/h3>\n

Several organizations are implementing large-scale machine learning models, but the data required to train these models raises concerns. There are two main issues: on one hand, a single organization may not have enough data to train large models, or may lack data on critical cases. On the other hand, even within the same organization, data may be sensitive to business operations, preventing it from being pooled into a central repository for model training.<\/p>\n

Solution<\/h3>\n

The FLaaS architecture provides for each data provider to participate by running a peer service, registering with a coordination layer at startup and describing the data available for training. The coordination layer is expected to manage the availability of each peer’s data and resources, providing an interface for users to submit training jobs and matching job requirements to a list of compatible peers. This will be done in compliance with current data policy and usage limitations.<\/p>\n

Benefits<\/h3>\n

The framework resulting from this project will allow us to overcome the main challenges related to federated learning: it will no longer be necessary to build a new federation from scratch every time a new model is trained. In addition, thanks to collaboration among partners, the architecture will make larger datasets available for training. This will be done in compliance with current policies, making data use more secure.<\/p>\n

IFAB’s role<\/h3>\n

IFAB took part in the project in an operational capacity, contributing to the management and coordination of activities across the different development phases. This involvement covered both organizational aspects and liaison among the parties involved, ensuring consistency between the activities carried out and the project’s objectives. Through this contribution, IFAB helped ensure the project’s proper execution and the delivery of the expected results, making its management expertise available in support of the entire process.<\/p>\n

Partners<\/h3>\n