Enveil has released its new encrypted training solution, ZeroReveal ML Encrypted Training (ZMET).
The enterprise-ready product expands the boundaries of trusted computing by enabling the secure use of separate, decentralized datasets for encrypted federated learning and machine learning applications.
Designed to address specific customer pain points, ZMET allows organizations to train models in an encrypted capability while protecting the process of model development, the model itself, and the interests of all parties involved. The product expansion, an extension of Enveil’s Machine Learning Solution Suite, comes after the company announced $ 25 million in Series B funding.
The rise of the digital economy is driving a broader market to expand the global data silo and gain insights through the use, analysis and machine learning of secure and personal data. Enveil’s ZeroReveal solutions reinforce this digital transformation by changing the example of how and where companies can use data to unlock values.
A recent Gartner report “Innovation Insights for Federated Machine Learning”, which recognizes Enveil as a representative provider, highlights the momentum in this market: To create more accurate, safe and environmentally sustainable models “(March 2022).
“Today’s digital-first business landscape demands solutions that extend an organization’s reach without sacrificing privacy or security,” said Dr. Allison Ann Williams, founder and CEO of Enville. “By ensuring that models are trained safely – and the model itself and its associated results are encrypted – ZeroReveal allows machine learning companies to leverage ML so that they can safely gain insights from data sources across silos, jurisdictions or borders, even using highly sensitive models. Training information while doing. “
ZMET Privacy uses advances in enhanced technologies, such as Secure Multiparty Computing (SMPC), to train models in an encrypted capability. This encrypted training process enables secure federated learning, the model development process, the data used for training, as well as the interests and intentions of the parties involved.
Companies can confidently lift sensitive data and / or ML models without the risk of exposure, providing improved models that can be used to more accurately gain insights and provide value. Models can be trained using data sources across security domains and organizational boundaries without the risk of unwanted exposure.
“We are proud to be the first in our department to provide an encrypted training product with a concrete and verifiable security approach: ZMET is providing an unparalleled ability to gain insights from data without having to trust other parties when calculating,” said Dr. Ryan Carr, Chief Technology Officer of Enville. “These privacy-saving machine learning training capabilities are tailored to the needs of our customers, engineered to overcome barriers and add business and mission value to the use of ML and data science.”
At its core, ZeroReveal Machine Learning is a two-way, proxy layer software system that enables decentralized, distributed evaluation and training of encrypted machine learning models across multiple datasets. Enveil protects the content of the search, analytical, or machine learning model – and its associated results.
The company’s decentralized approach allows data to be used securely within entities and across organizational, jurisdictional and security boundaries, extending data utilities without the need to transfer or pool sensitive assets.
