endee-io/endee

Endee is an open-source vector database written in C++ and built for AI-powered search and retrieval. A vector database stores data as numerical representations (called vectors or embeddings) instead of plain text, which lets you search by meaning rather than exact keywords, for example, finding documents that are conceptually similar to a query even if they share no words. It is designed to handle up to one billion vectors on a single machine, making it suitable for large-scale AI applications. Endee supports dense vector search (meaning-based similarity), sparse search (term-precision matching), and hybrid search that combines both. You can also layer in payload filtering, which means narrowing results by structured metadata like categories, dates, or tags alongside the semantic search. Common use cases include building RAG pipelines (Retrieval-Augmented Generation, a technique where an AI assistant fetches relevant documents before answering), semantic search over documents or products, AI agent memory stores, and recommendation systems. The README specifically mentions compatibility with AI frameworks like LangChain, CrewAI, AutoGen, and LlamaIndex for agent memory use cases. The server runs locally or in Docker and exposes an HTTP API on port 8080. Builds are optimized for modern CPU instruction sets including AVX2, AVX512, NEON, and SVE2. A hosted cloud version is also available at endee.io. The project is licensed under Apache 2.0.