Homomorphic encryption technology company Cryptolab said on Tuesday its homomorphic encryption-based near-real-time vector search engine, enVector, has been officially listed on Google Cloud Marketplace.
EnVector supports near-real-time vector search without performance degradation or delays even when data remains encrypted.
In computing, a vector refers to unstructured data such as text, images and audio converted into numerical arrays and represented as coordinates that capture meaning. It is used in generative AI and retrieval-augmented generation (RAG) to find semantically similar data beyond simple keyword matching.
Existing vector search can have a security vulnerability in which sensitive data such as internal corporate documents or customer information may be exposed during analysis, but enVector supports calculating similarity between vectors and searching encrypted data.
Cheon Jeong-hee (천정희), Cryptolab's chief executive, said she hopes the listing on Google Cloud Marketplace will give more companies a chance to experience homomorphic encryption technology directly in real cloud environments and to consider and use secure AI environments without security concerns.
Hwang Chi-kyu
delight@d-today.co.kr