sobowalebukola/memcortex
📦 Open Source Projectsobowalebukola
A lightweight semantic memory layer for LLMs using Go, Weaviate, and open-source embeddings.
Memcortex acts as a middleware memory layer, enabling LLMs to maintain persistent state through semantic search. Built in Go, it is optimized for performance and ease of deployment via Docker. The architecture centers on a vector-first approach, utilizing Weaviate as the primary storage engine to handle high-dimensional embeddings. Key features include an efficient pipeline for embedding generation, structured vector retrieval, and a modular design that allows developers to swap out embedding models. This tool is particularly effective for RAG (Retrieval-Augmented Generation) workflows, where maintaining a coherent history or accessing a large knowledge base is critical for agentic performance. Its lightweight footprint makes it an ideal choice for developers looking to add 'brain' functionality to their AI agents without the overhead of heavy, proprietary memory platforms.
💡Highlights
- ├─Go-native memory architecture
- ├─Weaviate-powered vector storage
- └─Seamless RAG integration
🎯For
- ├─AI Engineers
- └─Backend Developers