Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Securing vector embeddings for Gen AI RAG applications with VectorX Db
Learn how vector inversion can expose data, why query logs leak information, and how VectorX DB enables encrypted search with client‑side encryption and reduced computations.
Vector databases used in Gen AI RAG applications are not secure. Vector Inversion techniques can be applied to recover large part of the data if the Vector databases is stolen. Then there are network logs that have traces of queries performed on vector databases that can contain sensitive data.
VectorX DB algorithm enables searching on encrypted data without decryption. And, it applies a unique approach to drastically reduce the number of computations required in order to achieve super fast decryption. It uses a client side encryption.
VectorX DB leverages Queryable Encryption and Hybrid Graph Memory for secure vector search.
Compose Email
Loading recent emails...