Erl VectorDB is a robust, high-performance vector database built on Erlang/OTP, meticulously designed to harness the power of the Actor model and OTP supervision trees. This architecture ensures exceptional fault tolerance, massive concurrency, and horizontal scalability for demanding AI/ML applications. It integrates a full Model Context Protocol (MCP) implementation, a REST API, and OAuth 2.1 authentication, making it a versatile solution for managing, searching, and securing vector embeddings in distributed environments, further enhanced by advanced vector compression and hot code reloading capabilities.
Key Features
01Distributed Clustering with Automatic Replication and Failover
02Vector Compression with Multiple Algorithms for Storage Efficiency
03Dynamic Store Management and Hot Code Reloading
04OTP-Native Architecture for Fault Tolerance and Concurrency
050 GitHub stars
06OAuth 2.1 Authentication and Comprehensive REST API