Custom Mem0
Manages persistent memories for AI agents and applications, supporting user personalization through the Model Context Protocol.
About
Custom Mem0 provides a robust, production-ready memory service designed to empower AI agents and applications with persistent memory capabilities. It fully integrates with the Model Context Protocol (MCP), offering a standardized interface for agents to store, retrieve, update, and delete memories. Developers can choose between Neo4j for rich, graph-based memory relationships or Qdrant for high-performance vector similarity search, with pgvector handling embeddings. The service is containerized with Docker for easy deployment, includes comprehensive health monitoring, and features robust backup and recovery strategies, making it suitable for both development and production environments requiring personalized and context-aware AI interactions.
Key Features
- Flexible backend storage: Neo4j (graph-based) or Qdrant (vector-based) with pgvector
- Model Context Protocol (MCP) integration for standardized access
- Persistent Memory Management for AI agents and applications
- Production-ready with Docker containerization, health checks, and backup/recovery
- Comprehensive API with tools for memory operations (add, search, update, delete)
- 0 GitHub stars
Use Cases
- Enabling AI agents to maintain long-term, personalized memory across interactions
- Building AI applications that require persistent storage and retrieval of user-specific data
- Integrating custom memory capabilities into AI development environments like VS Code Copilot via MCP
