MAGIC

February 1st, 2015

“Insight from data – without seeing the data.”

Magic1

MAGIC enables the generation of new insights from sensitive data in situations where the data can be owned by a third party, segregated by jurisdictional boundaries, or be offered to third parties. Instead of data being centralised before processing, we take the computation to the data, and safely aggregate the results, removing trust issues and ensuring privacy for all parties including data providers, data consumers and service providers.

Magic4

MAGIC provides support for various analytics tasks – classification, clustering, and regression, etc. Data may be split between organisations in many different ways:

  • Shared features where two organisations have the same type of data about different entities; and
  • Shared entities where two organisations have different data about the same entities.

 

People: Stephen Hardy (Technical contact), Richard Nock, Max Ott,Brian Thorne, Guillaume Smith, Mentari Djatmiko, Giorgio Patrini, Roksana Borelli,