Opaque Systems Launches Multi-Party Confidential AI and Analytics Platform for Secure, Collaborative Data Science

Opaque Systems, provider of secure, multi-party analytics and AI for Confidential Computing, is launching its flagship platform—the Opaque platform—to expand the use cases for Confidential Computing and empower data scientists’ use of existing skills to combat current data analytics challenges.

Powered by the minds of the MC2 open source project, the Opaque platform leverages Confidential Computing, which is supported by all major cloud vendors, to enhance data encryption for secure, reliable, and collaborative analytics. Trusted Execution Environments (TEEs) are built to encrypt data during computation, further isolating it from access, exposure, and threats, according to the vendor.

“Data collaboration on confidential data has not yet truly been possible. Confidential data is currently locked down in silos, and as a result, organizations can't fully derive value from it; the value that's inherent to this confidential data is going untapped. It's getting lost.,” said Rishabh Poddar, CEO and co-founder of Opaque Systems. “And this problem is getting compounded with existing and emerging privacy laws and regulations. With Opaque, this data can now be activated while still maintaining its privacy and confidentiality guarantees.”

Though formerly too complex to utilize, users can benefit from TEEs, or “enclaves,” via the Opaque platform’s multi-party confidential analytics and AI, encouraging frictionless analytics on TEEs, secure data sharing, and collaborative analytics that allows multiple parties to have access to only the data they own. 

“As a user, you should not have to be an expert in confidential computing, or an expert in the underlying security and cryptographic mechanisms; you should be able to continue using the same tools, APIs, and interfaces that you currently use, and let the platform abstract away all the complexities of confidential computing from the user. That is what Opaque is about,” explained Poddar. “We make it frictionless for data scientists to run big data analytics and train machine learning models on their confidential data. And all the while keeping the data secure.”

The Opaque platform perfects collocated system access, where enclaved code and data are inaccessible to other users or processes within the same system. This is the foundation for Opaque’s truly collaborative and secure analytical process, as several data owners can group and analyze their encrypted data in the cloud without compromising security. Data owned by an individual party will never be exposed to other data owners; regulatory compliance is adhered to across departments and across organizations within TEEs.

Rich, multi-tiered defense—including Intel Software Guard Extensions, secure enclaves, advanced cryptography, and policy enforcement—allows Opaque to ensure code integrity, data, and side-channel attack protection. Said protection lasts the entire computation lifecycle, from data upload, to analytics, to insights. Further, enclave clusters overcoming performance and scaling issues are automatically orchestrated.

“And this is where perhaps, a number of other approaches fall short, is that this needs to be highly scalable in performance, because we're talking about very large quantities of data,” said Poddar. “So, our platform is, as a result, designed to be highly scalable. You will really inherit the scalability of the underlying Big Data platform that's Opaque—that forms its core.”

To learn more about the Opaque Confidential AI and Analytics Platform, please visit