Welcome to Tensorcrypt

We empower organisations to securely share and analyse sensitive data using confidential computing. At our core, we use trusted execution environments to mitigate information exposure risks at all analysis stages, while respecting performance, trackability, and usability.

Learn More

Overview

Tensorcrypt uses a client-side secure enclave solution, which is resistant to a wide variety of privileged software adversaries. This serves as a tamper-resistant black box for allowing organisations to exchange and analyse sensitive data. In tandem, Tensorcrypt allows data stakeholders to set-up secure sessions via a web platform.

Tensorcrypt architecture

We recognise that the high barrier of entry prevents businesses from unlocking the potential of privacy enhancing technolgoies. Our mission is to democritise secure data analysis techniques using high peformance and easy-to-use APIs that would otherwise require expensive, hard-to-resource expertise. Tensorcrypt fits into existing data workflows, supporting R and Python, for sharing data based on a common format agreed by organisations.

Designed with non-experts in mind

We abstract away the technicalities of privacy-enhancing data analysis, such as key management, so you can focus on what really matters.

Data is under your control

Our on-prem nature maximises reliability while minimising turnaround. Unlike existing solutions, data owners do not need to transfer data to our servers.

Decentralised Confidential Computing

We provide access to groundbreaking confidential computing that allows data analysis of sensitive data between organisations without revealing their contents.

Ready to Go

No need for expert input: use our platform to access encrypted data analysis immediately.

Immediate Results

Interact with our platform using a range of data analysis environments, including Python, R, Microsoft Power BI, and Excel.

Cutting Edge

Leveraging recent advances in confidential computing, we help you securely conduct data analyses without exposing sensitive business data.

Services

Tensorcrypt's ready-built functions allow technical and non-technical data analysts to leverage state-of-the-art confidential computing for secure data analytics.

Time-Series Forecasting

Perform time-series forecasting, such as regression analyses and moving averages, on sensitive datasets between end-points.

Recommendation Systems

Access state-of-the-art recommendation system algorithms without exposing your privacy-sensitive business data.

Image Recognition

Perform object detection and image recognition without disclosing raw images in unprotected form.

Sentiment Analysis

Leverage text-based sentiment analysis algorithms while keeping your input data confidential.

Exploratory Data Analysis

Perform statistical analysis, conditional filtering, and more.

Clustering

Exploit unsupervised learning to cluster your datasets in a secure fashion.

Tensorcrypt allows data users to securely share and analyse sensitive datasets between organisational end-points, unlocking their value by mitigating data in-use, in-transit and at-rest information exposure risks.

Please contact us for further information or read our one-page summary.

Contact Us Commercial Summary

Team

Tensorcrypt is spearheaded by experts in confidential computing. Collectively, they have published over 50 peer-reviewed scientific publications in the area in international events.

Dr. Carlton Shepherd

Co-Founder

Dr. Carlton Shepherd holds a Ph.D. in Information Security from Royal Holloway, University of London. He is a Senior Research Fellow at Royal Holloway, before which he was a Data Scientist at Atom Bank, the UK's first mobile bank, and a Research Scientist at OneSpan, a NASDAQ-listed provider of cybersecurity products for leading financial institutions. Carlton is an expert in trusted and encrypted execution environments, complemented by experience in industrial R&D and using data to inform business-critical services.

Prof. Konstantinos Markantonakis

Co-Founder

Prof. Konstantinos Markantonakis is a Professor of Information Security in Royal Holloway University of London, and the Director of the Information Security Group Smart Card and IoT Security Centre. He obtained his B.Sc. (Lancaster University), M.Sc., Ph.D. (London) and his MBA in International Management from Royal Holloway, University of London. He is an expert in smart card security and applications, secure cryptographic protocol design, embedded systems security, and trusted execution environments.

Contact

Contact us directly using the email below, or sign up to our mailing list for updates about our release.

Our Address

carlton.shepherd@rhul.ac.uk