Obtaining richer insights from data while preserving privacy and security obligations
Data collected on devices can be very sensitive, from personal health information tracked by a wearable device, to industrial sensor data capturing mission-critical information. Insights from this data can also be very valuable, from personalised advice about health to remote diagnosis of critical infrastructure.
More data means better insights for decision making. However, providing data to and obtaining data from other organisations can introduce security, regulatory, ethics and reputational risks.
Traditional methods of solving these challenges involve anonymising and restricting access to data but this can significantly reduce the value and insights that could be derived. In addition, issues of trust and the nature of security obligations between different organisations can make it difficult for organisations to collaborate and share critical information.
Confidential Computing – Federated privacy-preserving analytics platform
Our Confidential Computing platform combines distributed machine learning with homomorphic encryption and secure multiparty computing to provide the ability to learn models across multiple datasets without any of the data leaving its secure source (i.e. without the need for organisations to disclose the data in its raw form). This keeps the source data both private and up-to-date.
The results from the encrypted calculations are identical to the results processed in the clear – there is no loss of accuracy due to the encryption process.
Our Confidential Computing platform enables the analysis of device data without disclosing the data to anyone. Statistical and analytical models can be generated across large numbers of devices, such as those used for anomaly detection and failure prediction.
This enables data driven outcomes for a single device, based on the statistical behaviour of all devices, while keeping the data confidential.
Our platform enables organisations to combine data across an industry, or from suppliers or agents, to obtain insights without learning any of the participant's data directly. Participants can:
- understand their position relative to their peers without disclosing their data to anyone
- develop better models of their customers through higher resolution mapping of the market
- allow encrypted queries over their data for compliance or governance purposes without exposing the data.
Privacy preserving data gathering for impactful insights
Confidential computing can be tailored to meet the specific needs of our clients.
Contact us to find out about using our confidential computing platform in your organisation.