Open source VENumML framework enables developers to build and deploy ML models on encrypted data
Vaultree, a leader in advanced data encryption, announced today the release and open-sourcing of its proprietary technology stack, VENum (Vaultree Encrypted Numbers). Designed to address the scalability challenges of existing Full Homomorphic Encryption (FHE) schemes, VENum includes VENumpy (Vaultree Encrypted Numpy), an internal FHE library that facilitates secure and scalable machine learning operations, as well as VENumML (VENum Machine Learning), a Python library powered by Vaultree’s own encryption scheme to combine the power of machine learning and FHE. This groundbreaking technology empowers both cryptography-literates and neophytes to perform advanced ML tasks securely, ensuring data privacy without compromise.
VENumML reduces barriers to encrypted ML capabilities by providing built-in, high-speed models for many applications. While extensive, disruptive processing is usually required to perform ML tasks on FHE data, this technology is designed to minimize performance impact. This versatile library supports multiple data formats, including images, tabular data, unstructured data, graphs and time series, allowing anyone to leverage the power of machine learning with minimal disruptions while maintaining privacy.
Vaultree is leading the way in encrypted AI/ML and secure data enablement as concerns around AI abound. A report by Deloitte this year found that 40% of professionals rank data privacy as a top concern when it comes to generative AI. Vaultree’s innovative, real-time platform uses FHE to maintain analytical capabilities while enabling complex machine learning on encrypted data, with built-in models that include linear models, transformers, PageRank and time series. With these novel and powerful tools, users can now use AI/ML in all environments while meeting privacy requirements.
VENum allows data-intensive industries under strict regulatory scrutiny to securely analyze sensitive data and remain compliant while harnessing the power of AI/ML. It has never been easier to build secure and private-by-design applications in sectors such as financial services, healthcare, government or any industry where privacy is critical.
Vaultree’s groundbreaking, quantum-resilient technology offers versatile applications among various industries. For example, accessing and leveraging protected information such as genomic data in healthcare or financial records in finance has always posed challenges due to privacy concerns. VENum overcomes these barriers, enabling the secure incorporation of genomic data to enhance drug-protein interaction models and advance research into targeted drugs. In finance, VENum allows for encrypted sensitive file searching and ranking, where organizations can employ encrypted machine learning to securely discover and request access to AES-encrypted documents, driving innovation and efficiency. Additionally, VENum facilitates major breakthroughs in federated learning, allowing healthcare institutions to pool patient data for better modeling and financial organizations to securely share encrypted data, resulting in significant gains such as fraud loss reduction.
As an open source machine learning framework, VENum fosters community engagement, innovation, transparency and trust. Developers, researchers and industry partners around the world now have access to a library of powerful tools to build and deploy ML models on encrypted data in a privacy-centric manner, advancing enterprise capabilities in secure AI applications and driving innovation.
Ryan Lasmaili, Co-Founder and CEO, Vaultree, said: “Our breakthrough framework gives industries—especially those that manage sensitive data, like finance and healthcare—the capability to analyze data securely. By open-sourcing VENumML, we are expanding the possibilities of secure AI for all developers, encouraging greater innovation and engagement. VENum makes encrypted ML a workable and scalable option for privacy-preserving intelligence and sets a new standard for secure data enablement.”
Explore AITechPark for the latest advancements in AI, IOT, Cybersecurity, AITech News, and insightful updates from industry experts!
