uTile

Technologies forming part of uTilePET

Secure Multiparty Computation: Secure multiparty computation makes use of additive secret sharing, which allows the segmentation of secret data into parts in such a way that none of the participating parties has the capability to reconstruct the original secret data

Federated Learning: Federated learning is an algorithmic solution which allows the training of ML models by sending copies of a model and carrying out the training where the data is stored

Private Set Intersection: It is used in cases of vertical partitioning. PSI is a cryptographic technique which allows the intersection to be found among various datasets without having to expose the data

Differential Privacy: This technique permits data to remain anonymous by deliberately injecting noise into the data set, so that all manner of statistical analyses of huge utility can be performed

It ensures that sensitive data is never exposed or transferred via departments, organisations or geographies.

The owners of the data do not have to entrust third parties with their data.

The data remain protected behind internal controls, whether on-premise or cloud, and any sensitive information remains private during computations.

An additional advantage of this new paradigm is that it opens the doors to cooperation between organisations from different fields, for example, public-private or inter-sector collaboration exploiting synergies.