Federated Learning for Artificial Intelligence Applications

Sherpa.ai is a Spanish deep-tech company pioneering the development of privacy-preserving AI solutions. Sherpa's federated learning platform allows organizations to train AI models collaboratively without ever sharing their data, thereby overcoming one of the main obstacles to AI adoption - data privacy. Around 75% of global business data remains unused in organizations due to privacy, regulatory, and competition barriers. Sherpa's platform can unlock learnings from this data to enable novel AI applications in sectors such as healthcare, energy, and cybersecurity.

What sets Sherpa.ai apart is its two-layer privacy architecture, combining federated learning with advanced Privacy-Enhancing Technologies (PETs) like Differential Privacy and Homomorphic Encryption. The platform keeps all data within its original environment and secures model updates using advanced mathematical methods, making it unique among current solutions.

Sherpa.ai is currently at TRL 6, with successful pilot projects and early license sales already completed. The EIC Accelerator project will support the platform’s further development to TRL 8, market readiness, and European scaling. With strong demand across regulated industries, validated customer use cases, and a highly experienced leadership team, Sherpa.ai is positioned to become a cornerstone of Europe’s AI sovereignty by unlocking siloed data without compromising privacy.