Blyss provides a Confidential AI platform that serves AI models using end-to-end encryption within secure enclaves. This architecture ensures data remains encrypted even during processing, preventing unauthorized access by anyone, including the provider. The platform offers verifiable security through browser enforcement and public logging of cryptographic attestations for applications like code copilots and semantic search.
Funding
$500K raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
AI models often require access to sensitive data for training and inference, creating a risk of data leaks and privacy breaches. Existing security measures may not fully prevent unauthorized access to data during processing, especially when using third-party AI platforms. This lack of robust data protection hinders the development and deployment of AI applications in privacy-sensitive domains.
Solution
Blyss provides a confidential AI platform that uses end-to-end encryption and secure enclaves to process data within a trusted execution environment. This approach ensures that data remains encrypted even during AI model training and inference, preventing unauthorized access by any party, including Blyss itself. By leveraging hardware-based security, Blyss enables organizations to develop and deploy AI-powered code copilots, semantic search, and personalized assistants without compromising data privacy. The platform offers browser-enforced security, verifying the server's confidentiality proof through a public Certificate Transparency log.
Target Audience
Blyss targets organizations that require the highest levels of data privacy and security when developing and deploying AI models, including those in regulated industries such as finance, healthcare, and government.
Features
- End-to-end encryption to protect data during AI processing
- Secure enclaves utilizing NVIDIA Hopper GPUs and AMD Genoa CPUs for trusted execution environments
- Public attestation logs for continuous verification of confidentiality guarantees
- Rust SDK for private data access built on homomorphic encryption
- Shuffle-and-perturb proxy for use with vector databases like Pinecone
- Support for code copilots, semantic search, and personalized assistants