Endee
ActiveDescription
A high-performance vector database designed to handle up to 1 billion vectors on a single node, delivering significant performance gains through optimized indexing and execution. Also available as a cloud service.
Key Features
- High-performance vector database built in C++ with optimized indexing for up to 1 billion vectors on a single node
- Hybrid retrieval combining dense vector search with sparse vectors and payload filtering
- CPU-targeted builds supporting AVX2, AVX512, NEON, and SVE2 for maximum hardware utilization
- Payload filtering for metadata-aware retrieval and application-specific query logic
- Backup APIs and operational logging for production-grade deployment and observability
- Flexible deployment via local builds, Docker images, and prebuilt registry images
Use Cases
💡 Serving as the retrieval layer for RAG pipelines in question answering and chat applications
💡 Providing long-term memory and context retrieval for AI agents built with LangChain or LlamaIndex
💡 Building semantic search experiences for documents, products, and knowledge bases
💡 Supporting recommendation and similarity matching systems across text and embeddings
💡 Enabling hybrid search workflows requiring both semantic understanding and term-level precision
Tags
Categories
Quick Start
1. Clone the repo and make scripts executable with `chmod +x ./install.sh ./run.sh`. 2. Build for your CPU: `./install.sh --release --avx2`. 3. Start the server with `./run.sh`. 4. The server listens on port 8080. Use the HTTP API to create indexes and serve retrieval queries.