MemMachine
ActiveDescription
Universal memory layer for AI Agents providing scalable, extensible, and interoperable memory storage and retrieval to streamline agent state management for autonomous systems.
Key Features
- Three-tier memory: episodic (graph), profile (SQL), and working memory
- LLM-agnostic design supporting OpenAI, Anthropic, Bedrock, Ollama
- Integrations with LangChain, LangGraph, CrewAI, LlamaIndex, and n8n
- Native MCP server for Claude Desktop and Cursor integration
- Self-hosted (Docker/Neo4j) or managed cloud platform
- Python SDK, TypeScript SDK, and RESTful API interfaces
Use Cases
💡 Building personalized AI assistants with long-term user memory
💡 CRM agents that recall client history and deal stages
💡 Healthcare navigators that track medical history and treatments
💡 Writing assistants that learn style guides and terminology
💡 Multi-session chatbots with persistent context across conversations
Categories
Quick Start
pip install memmachine-client, initialize MemMachineClient, create a project and memory instance, then use memory.add() and memory.search() for persistent agent memory.