
Specializes in privacy-preserving LLMs and federated learning solutions for enterprise generative AI.
Provides an open-source community and enterprise platform for federated learning, focusing on distributed training and deployment.

Germany · Startup
Develops the Flower framework, an open-source, unified approach to federated learning that works with any workload, ML framework, and training environment.
A community-driven organization building privacy-preserving AI technology, including PySyft for encrypted, privacy-preserving deep learning.
Offers a platform for creating collaborative data ecosystems using federated learning and privacy-preserving technologies.
Provides a distributed data science platform that allows algorithms to travel to the data rather than moving the data itself.
A biotech company that uses federated learning to train AI models on distributed patient data without the data leaving hospitals.
Provides data clean rooms powered by confidential computing to enable secure data collaboration and model training.
Developing foundation models for robotics (Project GR00T) and vision-language models like VILA.
Provides a privacy-preserving AI platform that enables federated learning for data privacy and regulatory compliance.
Developer of the Loihi neuromorphic research chip and Foveros 3D packaging technology.
Federated learning platforms enable machine learning model training across distributed devices or institutions without centralizing data, addressing privacy and regulatory constraints. These systems coordinate training by sending model updates (rather than raw data) from local devices to a central server, which aggregates updates to improve the global model while keeping sensitive data on local devices.
This innovation addresses critical privacy and regulatory challenges in AI, particularly in healthcare, finance, and other sectors where data cannot be easily shared due to privacy regulations, competitive concerns, or security requirements. By training models on distributed data without centralizing it, federated learning enables collaborative AI development while maintaining data privacy. The technology includes additional privacy protections like differential privacy and secure aggregation to further protect sensitive information. Companies like Google, NVIDIA, and various startups provide federated learning platforms.
The technology is essential for enabling AI in privacy-sensitive domains where data sharing is restricted, such as healthcare (where patient data must remain in hospitals), finance (where data sharing is regulated), and mobile devices (where user privacy is paramount). As privacy regulations tighten and concerns about data centralization grow, federated learning offers a pathway to collaborative AI that respects privacy and regulatory requirements. However, the technology faces challenges including communication efficiency, handling non-IID data distributions, and ensuring model quality without direct data access.